psi4/psi1
This commit does not belong to any branch on this repository, and may belong to a fork outside of the repository.
Folders and files
| Name | Name | Last commit message | Last commit date | |
|---|---|---|---|---|
 |  | |||
 |  | |||
 |  | |||
 |  | |||
 |  | |||
 |  | |||
 |  | |||
 |  | |||
 |  | |||
 |  | |||
 |  | |||
 |  | |||
 |  | |||
 |  | |||
 |  | |||
 |  | |||
 |  | |||
 |  | |||
 |  | |||
 |  | |||
 |  | |||
 |  | |||
 |  | |||
 |  | |||
 |  | |||
 |  | |||
 |  | |||
 |  | |||
 |  | |||
 |  | |||
 |  | |||
 |  | |||
 |  | |||
 |  | |||
 |  | |||
 |  | |||
 |  | |||
 |  | |||
 |  | |||
 |  | |||
 |  | |||
 |  | |||
 |  | |||
 |  | |||
 |  | |||
 |  | |||
 |  | |||
 |  | |||
 |  | |||
 |  | |||
 |  | |||
 |  | |||
 |  | |||
 |  | |||
 |  | |||
 |  | |||
 |  | |||
 |  | |||
 |  | |||
 |  | |||
 |  | |||
 |  | |||
 |  | |||
 |  | |||
 |  | |||
 |  | |||
 |  | |||
 |  | |||
 |  | |||
 |  | |||
 |  | |||
 |  | |||
 |  | |||
 |  | |||
 |  | |||
 |  | |||
 |  | |||
 |  | |||
 |  | |||
 |  | |||
 |  | |||
 |  | |||
 |  | |||
 |  | |||
 |  | |||
 |  | |||
 |  | |||
 |  | |||
 |  | |||
 |  | |||
 |  | |||
 |  | |||
 |  | |||
 |  | |||
 |  | |||
 |  | |||
 |  | |||
 |  | |||
 |  | |||
 |  | |||
Repository files navigation
PPPPPPP SSSSSS IIIIII
PPPPPPPP SSSSSSSS IIIIII
PP PP SS SS II
PP PP SS II
PP PP SSS II
PPPPPPPP SSSSS II
PPPPPPP SSSSS II
PP SSS II
PP SS II
PP SS SS II
PP SSSSSSSS II
PP SSSSSS IIIIII
USER MANUAL
(Ver 1.0, Feb 1989)
(Ver 1.1, Jul 1989)
Edited by
Yaoming Xie and Brian Yates
4
1
1
Table of Contents
=================
1. Introduction
2. Capabilities of PSI
3. REALLY USEFUL INFORMATION
4. Order of programs required to run a calculation
5. Detailed description of each program
6. Examples
Appendix I: List of source files
Appendix II: Index of files used by PSI
Appendix III: Execs for IBM VM/CMS operating system |
|
Appendix IV: Historical list of Schaefer-group members
Appendix V: Index for Chapter 5.
Sections of this manual (such as this sentence) which |
are flanked on the right hand side by a "|" are not for |
general release to the public. |
1
1 I. Introduction
=================
PSI is a state-of-the-art suite of computer programs for the ab
initio quantum mechanical prediction of molecular structure, molecular
spectra, molecular properties, and chemical reactivity. These programs
have been under development in the research group of Professor H. F.
Schaefer III since 1969, first at the University of California,
Berkeley, later at the University of Texas, Austin, and now at the
University of Georgia, Athens.
Any publications arising from use of this software package should
include the following citation:
PSI 1.0, 1989, PSITECH Inc., Watkinsville, Georgia, USA
1
1 II. Capabilities of PSI
=========================
a. Single point energies
---------------------
SCF
Restricted closed- and open-shell (NOT UHF) Hartree Fock wavefunctions
(including high-spin open shells, and open shell singlets)
Singlet excited states |
Closed shell two configuration (TC) SCF
Paired-excited (PE) MCSCF
CI (configuration interaction)
Graphical Unitary Group Approach (GUGA) CI with shape driven
algorithm for single and double excitations
Flexible DRT allows any set of reference wavefunctions and any
level of excitation
CC (coupled cluster)
Closed shell single and double excitations (CCSD)
Closed shell single, double and linearized triple excitations
(CCSDT-1)
b. Derivatives of the energy
-------------------------
SCF
First, second and third analytic derivatives
These are calculated in the AO basis. |
(Second derivatives may also be calculated in the molecular |
orbital (MO) basis.) |
CI
First derivatives (with SCF and TCSCF reference wavefunctions)
CC
First derivatives (with closed shell SCF reference wavefunctions)
Derivatives may be transformed from Cartesian coordinates to
internal coordinates and vice versa
Higher derivatives are available through finite displacements
c. Optimizations
-------------
Minima and transition structures
Cartesian coordinates or symmetrized internal coordinates may be
used
d. Properties
----------
Mulliken and Lowdin population analyses
Dipole moments, electric polarizabilities
Vibrational frequency analysis (in normal coordinates, simple
internal coordinates and symmetrized internal coordinates)
Anharmonic constants
Infrared and Raman intensities
e. Limitations
-----------
Most arrays are dimensioned to allow
50 atoms
120 unique shells, and
256 primitive gaussian functions
In practice, disk space and memory availability will probably result
in limits which are much more restrictive than these.
f. Possible future directions
--------------------------
- analytical derivatives (SCF, CI and CC) for F- and G-type basis
functions
- use of five pure D-type basis functions and seven pure F-type
functions
- general contracted basis functions
- improved convergence of SCF for open-shell wavefunctions
- CI second derivatives
- MCSCF and MCSCF-CI extensions
- restricted open-shell CCSD
- CC gradient improvements (increased use of symmetry, use of
frozen core and virtual orbitals)
1 III. Really Useful Information
===============================
STOP! DO NOT SKIP OVER THIS SECTION!
Now that we have your attention... PSI is a very flexible
collection of programs but it can also be intimidating to the first time
user. There are many choices and numerous options. Most of these
options you will not use (and may not even want to know about) in the
beginning. These pages describe how to do the most common types of
calculations. In addition, some recommendations on wavefunction
convergence and a short list of common errors are given.
1. Take a look now at the input deck shown at the end of this chapter.
Only minor modifications of this "super input deck" are required to run
energies and gradients for SCF, CI and CC wavefunctions. We suggest
you make a copy of an input deck like this for yourself, and then make
the following changes to it when you want to do calculations on some
new molecule.
2. Getting started
Every study begins with the specification of a filename, the atoms
in the molecule, the basis set associated with each set of atoms,
and the point group symmetry of the molecule. These are defined in
the FILES and INPUT section. As far as PSI is concerned, once these
parameters are specified we may choose to determine any type of
wavefunction and associated energy, or any derivative of that
wavefunction that is compatible with the capablities of PSI, simply
by changing some parameters in subsequent sections. To begin, then,
you should do the following:
In the # FILES ## section, change CH2O to your filename.
In the # INPUT ## section, change CNV to your symmetry point group.
Change the atoms and the Cartesian coordinates to correspond to your
molecule. Change the basis set (GET DZ, etc.) to your basis set (see
Chapter V for details).
Now run INPUT. A file called SLOFILE is generated. Edit this file
and check that you have the correct number of basis functions, total
number of atoms, and so on.
3. SCF energy
Only four of the many options are routinely used here, and those are
the ones that specify the wavefuntion convergence, the number of open
shells, the flag to tell PSI whether to use a previous vector to aid
in SCF convergence, and the maximum number of SCF iterations. Below
the options line, we must specify the orbital occupations by symmetry
type and include a set of coupling coefficients if the electronic
state is not closed shell. Finally, if open shell states or TCSCF is
being used, then one should consider changing the DAMP and DAM1
options from their standard values of 0.0 and 1.0 to aid in the SCF
convergence. (WARNING: Please read the section entitled "Hints on
Converging the SCF" in the detailed description of the SCF program
in Chapter V before attempting open shell or TCSCF calculations).
Thus, in the # SCF #### section, set the second option to an
appropriate value for the SCF convergence, set the third option (IOPEN)
according to your wavefunction (closed shell = 0, open shell = 1,
TCSCF = 2). Set the fourth option = 1 if you have a previous guess
vector (else = 0). Set the tenth option to a reasonable number for the
maximum number of SCF iterations. Put in the orbital occupation numbers.
Add coupling coefficients if you are not doing a closed shell molecule.
Change DAMP and DAM1 if necessary.
Now you can calculate the SCF energy. Check the output to make
sure you have the correct electronic state and that the SCF converged.
4. SCF optimization
With the molecular geometry, point group symmetry, atomic basis sets
and electronic states defined as in sections 2 and 3 above, the
additional modifications required to obtain the derivative of the
SCF energy with respect to nuclear coordinates for use in geometry
optimization are:
Modify the # DERIV ##, # INTCOS # and # GNEXTS # sections for your
molecule according to the descriptions in Chapter V.
Now you can run the SCF derivative and geometry optimization
sequence of programs. If you want to use the B-matrix program, set up
the file called BMAT as in the example at the end of this chapter.
5. SCF vibrational frequencies
In the # DERIV ## section, change FIRST to SECOND.
In the # NORMCO # section, change the third option to the number of
degrees of freedom of your molecule.
Now you can calculate second derivatives and vibrational
frequencies. If you want to use the INTDER program, set up a file
called INTDER1 as in the example at the end of this chapter.
6. CISD energy
To perform a standard CISD calculation, only the input in the DRT
section needs to be modified. For the example we have shown here, the
number of irreducible reps is 4 and the number of basis functions is 60.
These need to be changed for your molecule. You also need to specify
how many molecular orbitals are frozen, how many occupied, and how many
unoccupied in the CI treatment. Thus, with the molecular geometry,
point group symmetry, atomic basis sets and electronic states defined
as in sections 2 and 3 above, CISD energies may be obtained as follows:
In the # DRT #### section, modify the number of irreducible reps
and number of basis functions for your molecule. Put in the appropriate
orbital codes (see Chapter V).
Now you can calculate the CISD energy.
7. CI optimization
With the molecular geometry, point group symmetry, atomic basis sets
and electronic states defined as in sections 2 and 3 above, the
coordinates defined either in the # INTCOS # and # GNEXTS # sections
or in the BMAT file as in section 4 above, and the correct # DRT ####
input as in section 6 above, the additional modifications required
to optimize the geometry with a CISD wavefunction are:
In the # MASTER # section, change it to read GRSCF CI FIRST.
In the # DRT #### section, change any FZC to COR, and FZV to VIR.
Now you can run CI gradients and the geometry optimization programs.
8. Coupled Cluster energy
With the molecular geometry, point group symmetry, atomic basis sets
and electronic states defined as in sections 2 and 3 above, CCSD
energies may be obtained as follows:
In the # CCSD ### section, modify the number of core and virtual
orbitals for your molecule, and set the first option of the second
line of options to an appropriate value for the energy convergence
(7 in the example we have shown here).
Now you can calculate the CCSD energy (closed shell systems only).
9. CC optimization
With the molecular geometry, point group symmetry, atomic basis sets
and electronic states defined as in sections 2 and 3 above, the
coordinates defined either in the # INTCOS # and # GNEXTS # sections
or in the BMAT file as in section 4 above, and the modification to
# MASTER # described in section 7 above, additional changes needed to
perform a CCSD optimization are:
In the # CCSD ### section, set the numbers of core and virtual
orbitals to zero, and give an appropriate value for energy convergence.
In the # DRT #### section, set the input up as though the
molecule had no symmetry and no frozen orbitals. That is, set the
number of irreducible representations = 1, and then give the number of
DOC and UOC orbitals.
In the # LAGTR ## and # GRCPHF # sections, change the third options
to 1.
In the # ZCCSD ## section, give an appropriate value for the
convergence of the coupled-perturbed CCSD equations (9 in the example
we have shown here).
Now you can run a CCSD gradient and optimize the geometry.
See Chapter V for more details on the input required for all these
programs.
Recommendations on Wavefunction Convergence:
Possibly THE single greatest waste of time in the use of
ab-initio quantum chemistry programs is the over-convergence of the
wavefunction. If all you want to know is the energy at a single point,
converging the density matrix to twelve decimal places is a tremendous
waste of time, particularly for open-shell or TCSCF wavefunctions
which converge so slowly and require lots of I/O. Of course, the more
you plan to do with the SCF wavefunction, the more accuracy is needed.
Thus, if you plan to determine any derivatives, or transform the
MO's for use in a correlated energy calculation, then ten or more
decimal places may be required to achieve the desired accuracy in
the final result. On the other hand, in the initial stages of a
geometry optimization of a molecule less accuracy is required, and
the wise user will take this into account, and will increase the
wavefunction convergence as the structure nears equilibrium. These
comments are especially appropriate for investigations at correlated
levels of theory.
With this in mind, we recommend the following:
A) for single-point energies, or initial stages of geometry
optimizations, converge the SCF to 10**-8, or 10**-9.
B) for single-point CISD, CCSD, or CCSDT-1 energies, or during
initial stages of geometry optimization, convergence of the
correlated density matrix to 10**-6, or 10**-7 is sufficient.
C) for optimized geometries, converge the SCF to 10**-10, and
the correlated density matrix to 10**-8.
D) If vibrational frequencies will be computed, converge the
SCF to 10**-10,or 10**-11, and all correlated energy
density matrices to 10**-9, or 10**-10.
List of common errors:
- wrong SCF energy
Solution: check the electronic state. Modify the orbital occupation
numbers if necessary. Maybe reorder the eigenvectors.
- wrong coordinates in BMAT file
Solution: check coordinates carefully. Make sure they transform
correctly under the various symmetry operations. Only use the
totally symmetric ones for optimizations.
- geometry in FILE30 does not match the specified symmetry
Solution: the geometry updating procedure has probably gone wrong.
Check it carefully. Make sure the coordinates in the BMAT file are
correct.
We wish you success in your calculations!
1 Sample input deck (belongs in a file called INPUT)
# FILES ################################################################
CH2O
# INPUT ################################################################
Input for CH2O, singlet, C2v symmetry, DZ + 2P basis set
CNV 2
0 0 0
CARBON 6.0 0.0 0.0 0.0
GET DZ
7 D 1
1 1.5 1.0
8 D 1
1 0.35 1.0
OXYGEN 8.0 0.0 0.0 2.283
GET DZ
7 D 1
1 1.5 1.0
8 D 1
1 0.35 1.0
HYDROGEN 1.0 0.0 1.792 -1.111
GET DZ
3 P 1
1 1.4 1.0
4 P 1
1 0.25 1.0
# SCF ##################################################################
SCF input for formaldehyde, singlet, C2v symmetry
0 11 0 0 0 0 0 0 0 100
5 0
0 0
1 0
2 0
0.0 1.0
# TFOCK ################################################################
GRSCF CI FIRST
0
# DERIV ################################################################
CLSCF SCF FIRST
0
# MASTER ###############################################################
CLSCF SCF SECOND
0
# DIPDER ###############################################################
0
# CPHFAO ###############################################################
0 0 0
# NORMCO ###############################################################
0 0 1 0
# PROPER ###############################################################
0 0 0
# BONDEX ###############################################################
0 0 0
# INTCOS ###############################################################
3 2 0 0 0 0 0 0 0 0
1 2
1 3
1 4
3 1 2
4 1 2
# GNEXTS ###############################################################
3 1 1 0 0 0 0
1 2 4
UP 0 0 3
1 1
1 2
2 3 4
# DRT ##################################################################
DRT for CH2O, singlet, C2v symmetry, DZ + 2P basis set
0 2 0
4 60 1
2FZC1 3DOC1 23UOC1 2FZV1
6UOC2
1DOC3 9UOC3
2DOC4 12UOC4
# CI ###################################################################
0 15 0 0 9 0 0 0 0
# LAGTR ################################################################
1 0 0
# NEWDER ###############################################################
CI FIRST 0
# GRCPHF ###############################################################
0 0 0
# CCSD #################################################################
CCSD input for CH2O, singlet, C2v symmetry, DZ + 2P basis set
0 0 0 0 0 0 0
7 30
2 2
0 0
0 0
0 0
CCSD
# ZCCSD ################################################################
CCSD input for CH2O, singlet, C2v symmetry, DZ + 2P basis set
9 30
NORM
# ZMAT #################################################################
1 6
2 1 1.208 8
3 1 1.116 2 121.8 1
4 1 1.116 2 121.8 3 180.0 0 1
0 0 0.0 0 0.0 0 0.0 0 0
# GEOMUP #
0.0000000000 0.0000000000 0.0000000000
0.0000000000 0.0000000000 2.2827898095
0.0000000000 -1.7923684095 -1.1113154976
0.0000000000 1.7923684095 -1.1113154976
# GEOML ##
------------------------------------------------------------------------
Example of the auxillary input file called BMAT:
BMAT Formaldehyde, singlet, C2v symmetry (bfy)
CARD 4
FMAT
EIGF 1
PRIN
C 6 0.000000000000 0.000000000000 0.000000000000
O 8 0.000000000000 0.000000000000 2.283000000000
H 1 0.000000000000 -1.792000000000 -1.111000000000
H 1 0.000000000000 1.792000000000 -1.111000000000
0.000000000000 0.000000000000 0.065684003900
0.000000000000 0.000000000000 -0.067492560300
0.000000000000 0.000423688200 -0.000375082000
0.000000000000 -0.000423688200 -0.000375082000
K 1.0 STRE 1 2
K 1.0 STRE 1 3
1.0 STRE 1 4
K 1.0 BEND 3 1 4
13.0
0.0 4.9
0.0 0.0 1.0
STOP
------------------------------------------------------------------------
Example of the auxillary input file called INTDER1:
# FILES ################################################################
CH2O
# INTDER ###############################################################
4 6 6 2 0 0 0 0 0 0 3 1
STRE 1 2
STRE 1 3
STRE 1 4
BEND 2 1 3
BEND 2 1 4
OUT 2 1 4 3
1 1 1.0
2 2 1.0 3 1.0
3 4 1.0 5 1.0
4 6 1.0
5 2 1.0 3 -1.0
6 4 1.0 5 -1.0
0
12.00000
15.99491
1.007825
1.007825
------------------------------------------------------------------------
1 IV. Order of programs required to run a calculation
======================================================
PSI is a package of separated programs, in contrast to some other
ab initio packages which consist of just one very large piece of code.
This means that, whereas with some packages one can run one program
to do everything, with the approach used here one needs to run a
sequence of programs in a particular order (using, for example, a
macro procedure). This segmentation of the code results in maximum
flexibility for developing new algorithms and applying them to difficult
problems. It is very easy to run an isolated part of the package if
various files have been kept from earlier stages of the calculation.
This chapter describes the different types of calculations that can be
performed with PSI, and the particular subset of programs required
for each calculation.
a. Preliminary
-----------
There are three programs (INPUT, ZMAT and GEOMIU) that are used to
set up the basis set and geometry information.
INPUT This is the first program that must be run when a set
of calculations is begun on a molecule. This creates a
binary file called FILE30 (see Apendices for a
discussion of filenaming conventions) which is the
starting point for all the other programs.
If it is desired to change either the basis set or the
symmetry of the molecule then INPUT must be run again.
Normally, if one were doing, say, a geometry
optimization, frequency calculation and single point
energies for a particular conformation with a particular
basis set then INPUT would only be run once at the very
beginning.
ZMAT
GEOMIU These two programs may be used at any point in the
calculation to update the geometry in the file called
FILE30.
Note that ZMAT does not actually alter FILE30;
it simply takes a set of internal coordinates (in
Pople-like Z-Matrix format) and writes the corresponding
Cartesian coordinates to the bottom of the file called
INPUT. (See Chapter V for more details.)
GEOMIU can be used in conjunction with the ZMAT program,
or with other programs that write Cartesian coordinates
to the INPUT file, or indeed with a set of Cartesian
coordinates that the user has typed in by hand, to
update the geometry in the binary file, FILE30.
In the ensuing sections, it is assumed that the program INPUT has
already been run and that a file called FILE30 exists.
b. Single point energies
---------------------
SCF (ground state) and TCSCF
To calculate the ground state Hartree-Fock SCF energy of a
molecule, one needs to run the following programs in the order
specified:
INTS
SCF
This sequence of programs may also be used to calculate the
closed shell TCSCF energy of a molecule by setting the
appropriate options in the input for the SCF program (see
Chapter V for more details).
SCF (excited state) |
|
For singlet excited electronic states of the same symmetry as the |
ground state, one needs to run SCFX instead of SCF. |
Thus the order of programs required is: |
|
INTS |
SCFX |
|
|
Paired-excited multi configuration SCF (PEMCSCF)
The order of programs required is:
INTS
GVBSCF
Configuration Interaction (CI)
The order of programs required is:
INTS
SCF
DRT
TRANS
CISORT
GUGACI
ONEPDM (optional)
This sequence of programs may also be used to calculate the
TCSCF CI energy of a molecule. First, the appropriate options in
the input for the SCF program need to be set up for a TCSCF
calculation (see Chapter V). Then, in the input for the DRT
program the two orbitals that change their occupancy in the two
SCF configurations should be labelled as special (SPE) and the
extra code 'GVB' should be included (see Chapter V and the
Examples).
Coupled Cluster (CC)
The order of programs required is:
INTS
SCF
CCTRANS
NCCSRT
NCC9
c. First derivatives of the energy
-------------------------------
SCF
The order of programs required is:
INTS
SCF
DERIV
These codes may be used to calculate derivatives for SCF and
TCSCF wavefunctions.
For SCF excited state wavefunctions, one should use: |
|
INTS |
SCFX |
DERIV |
Paired-excited multi configuration SCF (PEMCSCF)
The order of programs required is:
INTS
GVBSCF
GVBDER
Configuration Interaction (CI)
The order of programs required is:
INTS
SCFTFK
MASTER
DRT
TRANS
CISORT
GUGACI
ONEPDM
TWOPDM
LAGTR
CIPROP
BONDEX
DERTFK
NGRCPHF (use CICPHF for TCSCF reference wavefunctions)
Coupled Cluster (CC)
The order of programs required is:
INTS
SCFTFK
CCTRANS
NCCSRT
NCC9
NZCCSD
CCDMAT3
MASTER
DRT
CCTODRT
LAGTR
CIPROP
BONDEX
DERTFK
NGRCPHF
d. SCF second derivatives of the energy
------------------------------------
In calculating the analytic second derivatives, one has the
choice of using the supermatrix (PK-file) formulation or not.
Use of the supermatrix is normally the desired method as it
speeds up the solution of the coupled perturbed Hartree-Fock (CPHF)
equations, however it does require some extra disk space.
The order of programs required is:
AO basis without supermatrix
INTS
SCF
(or SCFX for excited states) |
MASTER
DERIV
(DIPDER)
CPCLAO or CPGRAO or CPTCAO
(NORMCO)
(PROPER)
(BONDEX)
('CL' is used for closed shell wavefunctions,
'GR' is used for general restricted open shell wavefunctions,
'TC' is used for two configuration wavefunctions)
... and excited state wavefunctions) |
AO basis using supermatrix formulation
INTS
SCF
(or SCFX) |
MASTER
MAKE37
DERIV
(DIPDER)
CPCLAOS or CPGRAOS or CPTCAOS
(NORMCO)
(PROPER)
(BONDEX)
Another way of calculating analytic SCF second derivatives is to |
use the molecular orbital (MO) basis. This method employs an |
additional transformation but is usually more efficient for a |
small number of basis functions. |
The order of programs required is: |
|
MO basis |
|
INTS |
SCF (or SCFX for excited states) |
MASTER |
DERIV |
TRANSYY |
CPCLMO or CPGRMO or CPTCMO |
|
|
e. SCF third derivatives of the energy
-----------------------------------
The order of programs required is:
INTS
SCF
MASTER
DER3RD
(DIPDER)
CPCLAO or CPGRAO
CL3RD or GR3RD
(ANHARM)
('CL' is used for closed shell wavefunctions,
'GR' is used for general restricted open shell wavefunctions)
f. PEMCSCF second derivatives
--------------------------
The order of programs required is:
INTS
GVBSCF
MASTERPX
FORM37
GVBDER2
(DIPDERPX)
CPPXAO
(NORMCO)
g. PEMCSCF and TCSCF third derivatives
-----------------------------------
The order of programs required is:
PEMCSCF: TCSCF:
INTS INTS
GVBSCF GVBSCF
MASTERPX MASTERPX
FORM37 FORM37
NEW3RD NEW3RD
(DIPDERPX) (DIPDERPX)
CPPXAO CPTCAOX
PX3RD TC3RD
(ANHARM) (ANHARM)
h. Summary of available analytical derivatives
-------------------------------------------
derivatives
first second third
___________________________________________________
SCF CLSCF X X X
GRSCF X X X
TCSCF X X X
PEMCSCF X X X
CI CLSCF X
GRSCF X
TCSCF X
CC CLSCF X
i. Geometry optimizations
----------------------
To perform an optimization in Cartesian or internal coordinates,
the order of programs required is:
+->SCF, CI or CC gradient (see section c.)
| INTCOS
| GNEXTS or NEWTON
| GUESSSCF or GUESSCI (optional) |
| |
+----+
To perform an optimization in symmetrized internal coordinates, the
order of programs required is:
+->SCF, CI or CC gradient (see section c.)
| BMWRTA
| BMATIN6
| GEOMIU
| GUESSSCF or GUESSCI (optional) |
| |
+----+
Each of these methods basically consist of a loop through gradient,
optimization, and updating routines. Because of the separated
nature of the programs, there is no easy way at the moment to
automatically stop the optimization when a certain threshold has
been attained. Therefore, user interaction is usually required to
determine if convergence has been reached.
WARNING: Use of the programs GUESSSCF and GUESSCI is only permitted |
for experienced PSI users! In general, it is much safer |
to do one cycle of an optimization and then carefully |
check the energy, the gradients, the choice of coordinates, |
and the geometry update, before sending the programs off |
into the wilderness. |
|
Some extra notes on geometry optimizations and the use of symmetry
are given in Chapter V under the description of the BMATIN6
program.
j. SCF dipole moment derivatives
-----------------------------
(These are usually calculated in the SCF second derivative sequence.)
The order of programs required is:
INTS
SCF
(or SCFX) |
MASTER
(MAKE37)
DERIV
DIPDER
CPCLAO(S) or CPGRAO(S) or CPTCAO(S)
k. SCF polarizability derivatives
------------------------------
The order of programs required is:
INTS
SCF
MASTER
(MAKE37)
DERIV
(DIPDER)
CPCLAO(S)
or CPGRAO(S) |
RAMANC
or RAMANG ('G' for general restricted open shells) |
RAMINT
l. Properties
----------
There are three population analysis programs available:
PROPER Mulliken population analysis (SCF)
CIPROP Mulliken population analysis (CI and CC)
BONDEX Bond orders and valencies (Mulliken and Lowdin)
For SCF wavefunctions, PROPER and BONDEX can be run after
completing section b.
For CI and CC wavefunctions, CIPROP and BONDEX should be
incorporated into the CI and CC gradient sequence as shown in
section c.
The following programs can be run after completing sections d., f.,
j. and k. to obtain a vibrational analysis, and infrared and Raman
intensities:
NORMCO Cartesian coordinates
INTDER internal coordinates (no Raman intensities)
m. Calculating frequencies and derivatives from finite displacements
-----------------------------------------------------------------
Numerical second derivatives can be obtained through finite
displacement of analytic first derivatives. In a similar way,
numerical third derivatives can be obtained from analytic second
derivatives, and so on.
Two very common uses of these procedures are a) to calculate CI or
CC frequencies, and b) to obtain SCF fourth derivatives for an
anharmonic analysis.
The user may choose to perform the finite displacements in either
Cartesian or internal coordinates. These methods are described
separately below.
Cartesian coordinates
*********************
i) To carry out finite displacements of first derivatives in
Cartesian coordinates, proceed as follows:
A. Set up the reference geometry in Cartesian coordinates at the
bottom of the file called INPUT using the format appropriate
for the GEOMIU program.
Generate the displaced geometries by hand by adding and
subtracting values to the 3N Cartesian coordinates. Do this
only for the symmetry unique coordinates. (In the case of
H2O, there will be seven displacements. No out-of-plane
displacements are required. See the Examples.)
A recommended value for the displacement is 0.01 bohr.
The displaced geometries will not be all of the same
symmetry. They should be grouped according to their point
group and a different INPUT deck and corresponding FILE30
generated for each symmetry. (In the case of H2O, two
displacements will be C2v symmetry and the remaining five
will be Cs symmetry.)
B. Loop over the following programs n times, where n is the
number of displacements.
+->GEOMIU Updates FILE30 to the displaced geometry
| |
| SCF, CI or CC first derivatives
| |
+----+
The derivatives (in Cartesian coordinates) are accumulated
in FILE11. The geometry at the top of FILE11 should
correspond to the original (undisplaced) geometry.
(This loop will usually have to be performed separately for
each symmetry group of displacements (for example, C2v and Cs
in the case of H2O) and the resulting FILE11s concatenated.)
C. Run either VIBLRG or FORM15
This will generate a file called FILE15 containing the second
derivatives in Cartesian coordinates. The input for these
programs is described in Chapter V. With FILE15 in place,
vibrational analysis using either NORMCO or INTDER may be
performed (see section l.).
ii) The procedure for obtaining third and fourth derivatives from
finite displacements of analytic second and third derivatives is
similar to the above.
+->GEOMIU
| |
| SCF 2nd or 3rd Generates Cartesian coordinate
| | derivatives derivatives (FILE11, FILE15 and
| | FILE20)
| |
| Copy FILE15 to the end of TOTAL15
| Copy FILE20 to the end of TOTAL20
| |
+----+
After modifying TOTAL15 and TOTAL20 according to the input
descriptions in Chapter V,
run WRIT20 to obtain 3rd derivatives
run WRIT24 to obtain 4th derivatives
iii)To obtain dipole moment derivatives in Cartesian coordinates
from finite displacements of dipole moments,
run WRIT17
Internal coordinates
********************
i) The procedure for obtaining second derivatives from finite
displacements of first derivatives in internal coordinates is
as follows:
(this is a little complicated, so find yourself a nice quiet
spot where you won't be disturbed, take a deep breath,
and read on...)
A. Set up the geometry and internal coordinates in the file
called BMAT. (Note that you will need all 3N-6 coordinates
if you wish to calculate all the frequencies.)
Generate the displacements with the BMATIN6 program (see the
input description in Chapter V for more details).
Recommended values for the displacements are 0.005 Angstrom
for stretches and 0.01 radian for angles.
You will need +ve and -ve displacements for those coordinates
belonging to the totally symmetric irreducible representation
of the point group, and only +ve displacements for all the
rest (for the asymmetric coordinates, the -ve displacements
can be generated by symmetry operations of the point group).
The displaced geometries in Cartesian coordinates are written
in sequence to the bottom of INPUT. These geometries should
be grouped according to their symmetry point group and a
different INPUT and FILE30 generated for each symmetry.
(For example, H2O has two symmetrized coordinates of A1
symmetry (the symmetric stretch and the bend) and one of B2
symmetry (the asymmetric stretch). A total of five
displacements are required: two for each of the two A1
coordinates, and one for the B2 coordinate. The A1
displacements will be C2v symmetry, the B2 displacement Cs
symmetry. Thus, two INPUTs and two FILE30s will be
required.)
B. Loop over the following programs n times, where n is the
number of displacements.
+->GEOMIU Updates FILE30 to the displaced geometry
| |
| SCF, CI or CC first derivatives
| |
+----+
The derivatives (in Cartesian coordinates) are accumulated
in FILE11. The geometry at the top of FILE11 should
correspond to the original (undisplaced) geometry.
(This loop will usually have to be performed separately for
each symmetry group of displacements (for example, C2v and Cs
in the case of H2O) and the resulting FILE11s concatenated.)
C. Set up the input for the INTDER program with the options
NDER=1, NEQ=1, NINV=0, NFREQ=0, IRINT=0, NVEC=0, MULTI=n+1,
where n is the number of displacements.
Run INTDER
This will generate a file called FILE12 containing all the
first derivatives in internal coordinates.
D. Rename FILE12 to FILE12A and modify according to the input
description for the INTDIF program in Chapter V (see the
subsection entitled "Information required in FILE12A").
Run INTDIF
This will generate a file called IDER containing the non-zero
first and second derivatives in internal coordinates (also
some diagonal third derivatives).
E. Finally, set up the input for the INTDER program with the
option NINV=2 and copy the second derivatives from the IDER
file into the appropriate place in the input (after the
atomic masses). (Also set the options NDER=2, NEQ=0,
NFREQ=1 or 3, MULTI=1.)
Run INTDER
This will generate a file called FILE15 containing the second
derivatives in Cartesian coordinates. The frequencies will
be in the output file called INTDERO.
ii) The procedure for obtaining third and fourth derivatives from
finite displacements of analytic second and third derivatives is
similar to the above. Since the second and third derivatives in
Cartesian coordinates are not accumulated automatically, one
solution is to run INTDER after each displacement and accumulate
the internal coordinate derivatives:
+->GEOMIU
| |
| SCF 2nd or 3rd Generates Cartesian coordinate
| | derivatives derivatives (FILE11, FILE15 and
| | FILE20)
| |
| INTDER Generates internal coordinate
| | derivatives (FILE12, FILE16 and
| | FILE21)
| |
| Copy FILE12 to the end of FILE12A
| Copy FILE16 to the end of FILE16A
| Copy FILE21 to the end of FILE21A
| |
+----+
(The input for INTDER should have the options NDER=2 or 3,
NEQ=1, NINV=0, MULTI=0.)
After modifying FILE12A, FILE16A and FILE21A according to the
input description for the INTDIF program, one should then
proceed with steps D. and E. above. (In step E., NDER should be
set to 3 or 4.)
iii)To obtain dipole moment derivatives in internal coordinates from
finite displacements of dipole moments, at present one must do
the calculation by hand. (Note: in order to obtain correct
infrared intensities, you must use the atomic masses in the BMAT
file when generating the displacements using the BMATIN6
program. This ensures that the Eckart conditions are satisfied
(see Wilson, Decius and Cross "Molecular Vibrations" (1955),
Section 11-1).)
For a symmetric coordinate, S, the derivative is given by
d mu mu +ve - mu -ve
---- = ----------------- (for x, y and z)
d S S +ve - S -ve
For an asymmetric coordinate, S, the derivative is given by
d mu mu +ve - mu origin
---- = -------------------- (for x, y and z)
d S S +ve - S origin
These derivatives should be placed in FILE18 or in the file
called INTDER1 according to the input description for INTDER.
The program INTDER should then be run to obtain infrared
intensities (NFREQ=1, IRINT=1).
n. Anharmonic constants
--------------------
After obtaining fourth derivatives of the energy, a second-order
perturbation anharmonic analysis may be carried out by running the
program
ANHARM
o. Extra program
-------------
READ30
is a program mainly of use to more advanced users. It is described
fully in Chapter V.
1 V. Detailed description of each program
==========================================
This chapter provides details about each of the programs available
in the PSI package. For each program, the description is broken into
four parts:
A. function of the program
B. references
C. files used by the program
D. description of the input required
In part C., the files used by each program are identified by their
generic filetype (or file extension). For example, FILE6 or FILE30.
The full names of the files would be (under IBM VM/CMS):
filename FILE6
filename FILE30
and so on. The only exception is 'BASIS DATA', which is the full name
of the file containing the standard basis sets used by the program
called INPUT.
In part D., the input is listed as follows:
flag_name
line_1 format
option_name option_value
. .
. .
. .
line_2 format
option_name option_value
. .
. .
. .
Frequent reference should be made to the Examples.
There are a number of places in this chapter where options are
listed that are not presently available. These are planned for use in
future, more advanced versions of the programs.
1 The sections in Chapter 5. are arranged in the following order:
1. Preliminary
2. Files
3. INPUT
4. ZMAT
5. GEOMIU
6. INTS
7. SCF
8. SCFTFK
9. DERIV
10. DRT
11. TRANS
12. CISORT
13. GUGACI
14. ONEPDM
15. TWOPDM
16. LAGTR
17. DERTFK
18. NGRCPHF
19. CICPHF
20. CCTRANS
21. NCCSRT
22. NCC9
23. NZCCSD
24. CCDMAT3
25. CCTODRT
26. MASTER
27. MAKE37
28. CPCLAO/CPCLAOS
29. CPGRAO/CPGRAOS
30. CPTCAO/CPTCAOS
31. DER3RD
32. CL3RD
33. GR3RD
34. GVBSCF
35. GVBDER/GVBDER2
36 MASTERPX
37. FORM37
38. CPPXAO
39. CPTCAOX
40. NEW3RD
41. PX3RD
42. TC3RD
43. DIPDERPX
44. INTCOS
45. GNEXTS
46. NEWTON
47. BMWRTA
48. BMATIN6
49. DIPDER
50. RAMANC
51. RAMINT
52. PROPER
53. CIPROP
54. BONDEX
55. NORMCO
56. INTDER
57. VIBLRG
58. FORM15
59. WRIT17
60. WRIT20
61. WRIT24
62. INTDIF
63. ANHARM
64. READ30
65. SCFX |
66. TRANSYY |
67. CPCLMO |
68. CPGRMO |
69. CPTCMO |
70. RAMANG |
71. GUESSSCF |
72. GUESSCI |
1 1. Preliminary
-----------
Unless otherwise stated, the input for all the programs is read
from a file called (under IBM VM/CMS)
filename INPUT
where 'filename' is something meaningful to the user.
The data in the INPUT file is separated into sections with each
section being delimited by a flag of the form
# name # (format A10)
For example, the input section for the program DERIV begins
with the line
# DERIV ##
(i.e. one '#', one space, the name, one space, then enough '#'s to
round it out to ten characters. You MUST use ten characters,
otherwise the flag will not be matched. You can type anything you
like beyond the tenth position.)
The various sections in the INPUT file may be in any order.
________________________________________________________________________
2. Files (for IBM VM/CMS only)
-----
Each program needs to know where to read and store its information
(in binary form). This is achieved through the use of the
# FILES ## section.
Input format:
# FILES ##
1. FORMAT(A8)
FNAME = filename ... This should be the same as 'filename'
in the previous section.
This input section is required for (almost) every program. The
binary files (FILE30 - FILE99) produced by each program will be
called (under IBM VM/CMS):
filename FILE30
filename FILE35
filename FILE36
and so on.
________________________________________________________________________
1 3. INPUT
-----
A. INPUT is a preliminary program which reads the input data for the
molecule (geometry, basis set, etc. ) and generates a working file
called FILE30 which is the real starting point of each calculation.
INPUT can handle a total of 50 atoms, 120 unique shells, and 360
primitive gaussian functions. INPUT limits the use of symmetry
point groups to D2h and its subgroups.
B. Main references:
STO basis sets:
W. J. Hehre, R. F. Stewart and J.A. Pople, J. Chem. Phys. 51
(1969) 2657.
W. J. Hehre, R. Ditchfield, R. F. Stewart and J.A. Pople, J. Chem.
Phys. 52 (1970) 2769.
DZ and TZ basis sets:
S. Huzinaga, J. Chem. Phys. 42 (1965) 1293.
T. H. Dunning, J. Chem. Phys. 53 (1970) 2823.
Also see:
R. Poirier, R. Kari and I. G. Csizmadia, "Handbook of Gaussian
Basis Sets" Phys. Sci. Data 24 (Elsevier, 1985)
and references therein.
C. Files required: INPUT (# INPUT ##)
BASIS DATA
Temporary files used: none
Files generated: SLOFILE
CHECK
FILE30
D. Input format:
# INPUT ##
1. FORMAT(A80)
TITLE anything descriptive for the job
(Used for print out only)
2. FORMAT(F10.0,4I5)
TIMLIM (not used)
NGEOM = 0 ... zero out information from old calculations
= 1 ... preserve information from old calculations
NPRINT = 0 ... normal printing
= 1 ... extra printing for basis set + symmetry
NORMF = 0 ... normalize the basis functions (keep zero)
= 1 ... no normalization
NORMP = 0 ... if contraction coefficients correspond to
normalized primitive functions. (keep zero)
= 1 ... if contraction coefficients correspond to
unnormalized primitive functions
Values of the contraction coefficients for unnormalized (and
normalized) primitives are printed out.
If NORMF=0, the molecular orbital coefficients of the
occupied orbitals are given in terms of normalized
contracted basis functions.
If NORMP=0, the contraction coefficients of the D,F,G-type
primitive functions that are input should be the ones
corresponding to the normalized D(XX), F(XXX) and G(XXXX)
primitives.
3. FORMAT(A5,I5)
GROUP = C1 C1 group
CS Cs group
CI Ci group
CN Cn group
CNH Cnh group
CNV Cnv group
DN Dn group
DNH Dnh group
NAXIS order of principal axis for Cn, Cnv, Cnh, Dn, Dnh
(for this program NAXIS = 2 only)
For linear molecules, point groups C2v or D2h should be used.
4. Orientation of local symmetry frame
Lines 4 and 5 may be used to specify nonstandard orientation
of symmetry elements. (This option is particularly useful
for doing finite difference calculations with displaced
geometries generated by BMATIN6.)
To select the default orientation of symmetry elements for any
group other than C1, leave line 4 blank and omit line 5.
(See below for a description of the default orientation.)
FORMAT (6F10.5)
X1
Y1
Z1
X2
Y2
Z2
Point 1 = (X1,Y1,Z1), Point 2 = (X2,Y2,Z2)
These must be two distinct points.
For C1 group, omit lines 4 and 5, go to line 6.
For Cs group, any two points in the symmetry plane.
For Ci group, point 1 = center of inversion (ignore point 2).
For all other groups, points 1 and 2 may be any two points on
the local Z axis.
5. Orientation of symmetry frame (continued)
FORMAT (3F10.5,A8)
X3
Y3
Z3
DIRECT directional parameter
Point 3 = (X3,Y3,Z3)
This must NOT be collinear with points 1 and 2 on line 4.
For C1 group, omit line 5.
For Cs group, point 3 is any point in the symmetry plane;
DIRECT is not used.
For Ci group, omit line 5.
For all other groups,
if DIRECT = 'PARALLEL' point 3 lies on the local X axis
(this is the default),
if DIRECT = 'NORMAL ' point 3 lies on the local Y axis.
.....omit line 5 if default option is elected on line 4.
Group Default orientation of symmetry elements with
respect to local X,Y,Z frame
C1 no symmetry (omit lines 4 and 5)
Cs reflection plane = (X,Y)
Ci center of inversion is origin, X=Y=Z=0
C2 2-fold axis = Z
C2h 2-fold axis = Z, sigma-h plane = (X,Y)
C2v 2-fold axis = Z, sigma-v plane = (X,Z)
D2 principle 2-fold axis = Z, 2-fold axis = X
D2h principle 2-fold axis = Z, 2-fold axis = X, sigma-h
plane = (X,Y)
6. FORMAT(3I5)
CHARGE (not used)
MULTIPLICITY (not used)
IUNIT = 0 ... input geometry in atomic units
= 1 ... input geometry in Angstrom (program will
change it into atomic units)
7. FORMAT (A8,2X,F5.1,3F20.10)
NAME atom name (used only for print out)
ZNUC atomic number
X x-coordinate of the atom
Y y-coordinate of the atom
Z z-coordinate of the atom
8. Basis set input
Use standard basis sets and/or any gaussian-type basis
functions input by hand.
S, P and D functions may be used.
...also F and G |
Currently, F and G functions are only available for single- |
point energy calculations (i.e. no analytic derivatives). |
The present version of the program uses all six D-type
functions.
... and all ten F-type functions and all fifteen G-type |
functions. |
(1) Standard basis sets:
FORMAT(A80)
One line in input, options are:
GET STO STO-3G basis set, for H-Ar
GET DZ DZ basis set, for H, B-F, Al-Cl
GET DZP DZP basis set, for H, B-F, Al-Cl
GET TZ TZ basis set, for H, B-F, Al-Cl
GET TZP TZP basis set, for H, B-F, Al-Cl
GET DUNNING RYDBERG 3S for B-F
GET DUNNING RYDBERG 3P for B-F
GET DUNNING RYDBERG 3D for B-F, Al-Cl
GET DUNNING RYDBERG 4S for B-F, Al-Cl
GET DUNNING RYDBERG 4P for B-F, Al-Cl
GET DUNNING RYDBERG 4D for B-F
GET DUNNING NEGATIVE ION 2P for B-F, Al-Cl
GET WACHTERS 14s11p6d -> 10s8p3d, for Sc-Zn
GET 321G 3-21G basis set, for H-Ar
GET 631G 6-31G basis set, for H-Ar
GET 6311G 6-311G basis set, for H-Ne
GET 631GST 6-31G* basis set, for H-Ar
GET 631PGS 6-31+G* basis set, for H-Ar
GET 6311PPGSS 6-311++G** basis set, for H-Ne
GET PLUSS diffuse S (Pople), for H-Ar
GET PLUSP diffuse P (Pople), for H-Ar
Note: any mixing and matching is possible in principle. The
only restriction is that the basis functions be grouped in
ascending angular momentum quantum number (i.e. all the S's
first, then all the P's, then the D's, and so on).
For example, 3-21+G for a heavy atom could be set up as:
GET 321GS
GET PLUSS
GET 321GP
GET PLUSP
Notes on standard basis sets:
DZ means 9s5p -> 4s2p for B-F
11s7p -> 6s4p for Al-Cl
TZ means 9s5p -> 5s3p for B-F (i.e. TZ in
11s7p -> 7s5p for Al-Cl valence only)
(2) Gaussian basis functions input by hand (coefficients and
exponents):
FORMAT(I5,1X,A4,I5)
ISHELL (not used, but must be > 0)
ITYPE = ' S' or ' K' for S shells
= ' P' for P shells
= ' L' for L shells
= ' D' or ' M' for D shells
= ' F' for F shells |
= ' G' for G shells |
IGAUSS = number of contracted primitives in this shell
9. If line 8(2) is specified, then:
FORMAT(I5,E15.9,2E20.10)
KDUM = primitive number in this shell
(for each shell, KDUM takes values 1,2,...,IGAUSS)
(used for print out only)
EX = gaussian exponential parameter of the primitive
function
C1 = contraction coefficient for S,P,D,F,G shells, and
for the S function of an L shell.
C2 = contraction coefficient for the P functions of
an L shell
.....repeat line 9 IGAUSS times, one for each primitive of the
shell
For the data for the next shell, return to line 8.
10. A blank line ends the list of shells centered on this atom.
Repeat lines 7-10 until all the symmetry unique atoms have been
listed.
11. A blank line ends the list of symmetry unique atoms.
________________________________________________________________________
1 4. ZMAT
----
A. ZMAT is a geometry transformation program that reads in simple
internal coordinates (bond lengths, bond angles, etc.) using a
Pople-like Z-matrix and converts these into Cartesian coordinates.
The Cartesian coordinates are appended to the bottom of the file
called INPUT in the format appropriate for the GEOMIU program.
B. References: none
C. Files required: INPUT (# ZMAT ###)
Temporary files used: none
Files updated: INPUT
Files generated: FILE6
CHECK
D. Input format:
# ZMAT ###
The rest of the input is free format (i.e. put space(s) between
numbers).
Only include as many of lines 1-4 as are needed to specify your
molecule (i.e. for a triatomic, only lines 1-3 are needed).
1. NUM(1), ANZ(1)
2. NUM(2), Z(2,1), BL(2), ANZ(2)
3. NUM(3), Z(3,1), BL(3), Z(3,2), ALP(3), ANZ(2)
4. NUM(I), Z(I,1), BL(I), Z(I,2), ALP(I), Z(I,3), BET(I),
Z(I,4), ANZ(I)
.....repeat line 4 as needed to complete the geometry definition.
5. 0 0 0.0 0 0.0 0 0.0 0 0 (terminates input)
NUM(I) is the number of this center (=I) (integer)
ANZ(I) is the atomic number of this center. If ANZ=0, the
center is treated as a dummy atom. (integer)
Z(I,n) are used to define the internal coordinates
(integers)
BL(I) is the bond length between NUM(I) and Z(I,1) (real)
ALP(I) is the angle NUM(I) - Z(I,1) - Z(I,2) (real)
BET(I) If Z(I,4) = 0, then BET(I) is the dihedral angle
NUM(I) - Z(I,1) - Z(I,2) - Z(I,3)
If Z(I,4) = 1, then BET(I) is the bond angle
NUM(I) - Z(I,1) - Z(I,3) (real)
The first center is always placed at the origin of the
Cartesian coordinate system.
The second center is always placed along the positive Z-axis.
The third center is always placed in the X-Z plane (with
positive X-coordinate).
The dummy atoms are removed from the final Cartesian geometry
before it is written to the INPUT file.
________________________________________________________________________
1 5. GEOMIU
------
A. GEOMIU updates the geometry in FILE30. It searches for the
first occurrence of '# GEOMUP #' in the file called INPUT and reads
in the new coordinates. Then it writes these to FILE30 and
changes # GEOMUP # to # GEOM # in INPUT.
B. References: none
C. Files required: INPUT (# GEOMUP #)
FILE30
Temporary files used: none
Files updated: INPUT
FILE30
Files generated: FILE6
CHECK
D. Input format:
# GEOMUP ##
1. FORMAT(3F20.10)
COORD(1,I), COORD(2,I), COORD(3,I) X,Y,Z coordinates
.....repeat this line until all the atoms are listed
________________________________________________________________________
1 6. INTS
----
A. INTS calculates integrals in terms of symmetry-adapted atomic
orbitals (SO).
B. References:
Rys-polynomial method and integrals over gaussian basis functions:
H.F.King, and M.Dupuis, J. Comp. Phys. 21 (1976) 144.
M. Dupuis, J. Rys, and H. F. King, J. Chem. Phys. 65 (1976) 111.
J. Rys, M. Dupuis, and H. F. King, J. Comp. Chem. 4 (1983) 154.
Pitzer's method:
R. Pitzer, J. Chem. Phys. 58 (1973) 3111.
C. Files required: INPUT
FILE30
Temporary files used: none
Files updated: FILE30
Files generated: CHECK
FILE6
FILE34 one- and two-electron integrals
D. Input required: none
________________________________________________________________________
1 7. SCF
---
A. SCF carries out the iterative procedure to solve the
Hartree-Fock equations.
Note:
Since these programs are restricted to D2h symmetry and its subgroups,
and the orbital occupations are required to be integers, certain pure
angular momentum states derived from partial occupation of degenerate
orbitals cannot be obtained with the present codes. For example, the
2PIu (doublet PI u) state of linear O-N-O derived from the lowest energy
linear (pi u)1 configuration may only be computed as the 2B2u (doublet
B 2u) or 2B3u (doublet B 3u) component of the 2PIu (doublet PI u) state,
and the resulting spatial wavefunction will not have PI symmetry. In a
certain sense, however, this is desirable, as the energy will be a
continuous function of the bending angle. Calculating the energy of
bent configurations as 2B2u (doublet B 2u) or 2B3u (doublet B 3u) and
doing a pure 2PIu (doublet PI u) state at linear geometries results in
a pronounced discontinuity.
For the most part, triplet states resulting from double occupation
of a doubly degenerate orbital, such as the 3A2 (triplet A 2) state
resulting from the (e')2 or (e")2 configurations in D3h symmetry, or
the 3SIGMAg (triplet SIGMA g) state of a (pi g)2 or (pi u)2 configuration
in Dinfh (D infinity h) symmetry, will have the proper spatial symetry.
The singlet states resulting from these same electronic configurations
are inherently multiconfiguration and, as such, are not well represented
by single configuration wavefunctions.
Virial theorem:
In the present version of the program, the number printed out
as the "VIRIAL THEOREM" is incorrectly calculated and should be
ignored. Don't be put off by this; everything else is OK!
B. References:
PK-file method:
R. C. Raffenetti, Chem. Phys. Lett. 20 (1973) 335.
Molecular symmetry and closed shell HF calculations:
M.Dupuis, and H.F.King, Int. J. Quant. Chem. 11 (1977) 613.
DIIS for closed shell:
P. Pulay, Chem. Phys. Lett. 73 (1980) 393.
P. Pulay, J. Comp. Chem. 3 (1982) 556.
Coupling coefficients (alpha and beta) for open shell:
C. C. J. Roothaan, Rev. Mod. Phys. 32 (1960) 179.
Damping:
D. R. Hartree, "The Calculation of Atomic Structures" (Wiley: New
York) 1957.
M. C. Zerner and M. Hehenberger, Chem. Phys. Lett. 62 (1979) 550.
Level shifting:
V. R. Saunders and I. H. Hillier, Int. J. Quant. Chem. 7 (1973)
699.
C. Files required: INPUT (# SCF ####)
FILE30
FILE34
Temporary files used: FILE92
Files updated: FILE30 MO coefficients
Files generated: CHECK
FILE6
D. Input format:
# SCF ####
1. FORMAT(A80)
ALABEL title for SCF output (free field)
(for print out only)
2. FORMAT(14I5)
(1) IPRCT >= 0 ... number of iterations before extrapolation
< 0 ... alternative extrapolation method
(IPOPLE = .TRUE.)
(2) ISCF = 0 ... convergence on density matrix = 10**-5
n ... convergence on density matrix = 10**-n
(n=9 recommended for single point energies,
n=10 to 12 recommended for derivatives)
(3) IOPEN = 0 ... closed shell
= 1 ... open shell
= 2 ... TCSCF
(4) INFLG = 0 ... no initial guess for wave function
= 1 ... use last result
(see "Hints on converging the SCF" below)
(5) IVECT (keep zero)
(6) IPUNCH > 0 ... read in eigenvectors from FILE30 and reorder
them (see format of line 3)
(7) PRINT print option
(8) IDIIS iteration to begin using DIIS
(closed shell only)
(9) ISAVE (keep zero)
(10) ITRAS maximum number of iterations (default 40)
(May need 100-200 for open shell and TCSCF
wavefunctions if INFLG=0)
(11) MAXNO number of buffers desired (keep zero)
(12) ISTO threshold for elimination of basis functions
10**-ISTO (default is 10**-20)
(13) MICMX (not used)
(14) NCOR (not used)
3. If IPUNCH > 0, then:
FORMAT(14I5)
IORDER(II) for each symmetry irreducible representation,
list the new order of MO's. Begin each
irreducible representation on a new line.
All the irreducible representations should be listed, even
though for some of them there is no change in the MO ordering.
4. FORMAT(2I5)
NC(L) the number of doubly-occupied MO's for one
irreducible representation
NO(L) the number of singly-occupied MO's for one
irreducible representation
.....repeat this line for each irreducible representation
For TCSCF, NC(L) is the number of occupied MO's that do not
change their occupancy in the two configurations.
NO(L) = 1 for the two irreducible representations
containing a special orbital and zero otherwise.
The irreducible representations are ordered according to
Cotton's numbering. (i.e. 1 2 3 4 5 6 7 8
D2h Ag B1g B2g B3g Au B1u B2u B3u
D2 A B1 B2 B3
C2v A1 A2 B1 B2
C2h Ag Bg Au Bu
C2 A B
Ci Ag Au
Cs A' A" )
It is important to sit down and work out the desired
electronic configuration of the system being studied.
Obviously, the electronic state the user calculates will be
determined by the orbital occupancies given in this section.
5. If IOPEN is not equal to 0:
FORMAT(2F20.10)
ALPHA(I) open shell coupling coefficient (alpha)
BETA(I) open shell coupling coefficient (beta)
.....repeat this line MM*(MM+1)/2 times, where MM is the number of
symmetry irreducible representations containing singly-
occupied MO's
Examples:
for doublet:
0.0 -1.0
for triplet (with open shells of different symmetry):
0.0 -1.0
0.0 -1.0
0.0 -1.0
for triplet (with open shells of the same symmetry):
0.0 -1.0
for open-shell singlet:
0.0 -1.0
0.0 3.0
0.0 -1.0
for TCSCF: (constants supplied by program for TCSCF but a
dummy set still needed)
0.0 0.0
0.0 -1.0
0.0 0.0
for C1 symmetry, only one set of coupling coefficients is
possible (i.e. it is not possible to do open-shell
singlets or TCSCF in C1 symmetry with this program):
0.0 -1.0
for high-spin open-shell wavefunctions, the values of alpha
and beta are always 0.0 and -1.0, respectively.
6. FORMAT(2F20.10)
DAMP damping factor
DAM1 level shift parameter
(if > 0 and closed shell, DAM1=0.1)
Hints on converging the SCF:
INLFG option: For difficult open shell cases, it is recommended
that an appropriate closed shell calculation be run first (add or
remove an extra electron) and that this SCF vector then be used as
a guess (INFLG = 1) for the desired open shell wavefunction. For
TCSCF cases, it is always wise to run a closed shell (or perhaps
the appropriate triplet) SCF first and then use this as a guess for
the TCSCF.
Level shifting: For open shell systems, a level shift value of
0.5 to 3.0 is recommended. Start with a high value (2.0 - 3.0)
for the first SCF calculation and then reduce it (to 0.5 - 1.0)
for subsequent runs which use a converged SCF vector as the
starting point.
________________________________________________________________________
1 8. SCFTFK
------
A. SCFTFK performs a regular SCF calculation and then rotates the
molecular orbitals in order for it to be followed by a correlated
derivative calculation.
B. References: see SCF
C. Files required: INPUT (# SCF #### and # TFOCK ##)
FILE30
FILE34
Temporary files used: FILE92
Files updated: FILE30 MO coefficients (after rotation)
Files generated: CHECK
FILE6
FILE47
FILE49
D. Input format:
# SCF #### see the description for the SCF program
# TFOCK ##
1. FORMAT(3(A8,2X))
CALTYP = GRSCF ... for CI or CC gradients
Always use GRSCF for CI and CC gradients
even if a closed shell SCF reference
wavefunction is employed.
= TCSCF ... for TCSCF-CI gradients
CITYP = CI ... for CI gradients
= GVBCI ... for TCSCF-CI gradients
= MCSCF ... * not available at present
DERTYP = FIRST ... for first derivative (default)
2. FORMAT(I5)
IPRINT = 0 ... minimum printing
= 1-63 ... more printing (add powers of two)
________________________________________________________________________
1 9. DERIV
-----
A. DERIV calculates the AO derivative integrals up to second order for
SCF wavefunctions.
B. References:
P. Pulay, Mol. Phys. 17 (1969) 197, 204.
R. Moccia, Chem. Phys. Lett. 5 (1970) 260.
J. A. Pople, R. Krishnan, H. B. Schlegel and J. S. Binkley,
Int. J. Quant. Chem. Symp. S13 (1979) 225.
P. Pulay, J. Chem. Phys. 78 (1983) 5043.
H. F. Schaefer and Y. Yamaguchi, J. Mol. Struct. 135 (1986) 369.
Y. Osamura, Y. Yamaguchi, and H. F. Schaefer, Chem. Phys. 103
(1986) 227.
C. Files required: INPUT (# DERIV ##)
FILE30
FILE40 (if DERTYP = SECOND)
Temporary files used: none
Files updated: FILE30
Files generated: CHECK
FILE6
FILE11
FILE42 (if DERTYP = SECOND)
D. Input format:
# DERIV ##
1. FORMAT(3(A8,2X))
SCFTYP = CLSCF ... for closed shell SCF
= GRSCF ... for open shell SCF
= TCSCF ... for TCSCF
and excited state SCF |
= MCSCF * not available at present
CITYP = SCF ... for SCF derivatives
DERTYP order of derivative
= FIRST
= SECOND
2. FORMAT(I5)
IPRINT = 0 ... normal printing
= 1-4 .. more output
3. If SCFTYP = GRSCF:
FORMAT(A8,2X,I5)
OPTYPE = OPENTYPE
NUNIQ = 1 ... for doublet, triplet or any high-spin open
shell
= 2 ... for open-shell singlet
Lines 3 and 4 are NOT needed for TCSCF
4. If SCFTYP = GRSCF:
FORMAT(F10.5,10I5)
GOCC(I) occupation in Ith shell type - usually 1.0
(i.e. one electron in each orbital)
LL number of open-shells in Ith shell type
MOPN(I,J),J=1,LL list of numbers identifying the open-shells
in Ith shell type.
Usually MOPN(I,J)=1,2,...,LL
.....repeat this line NUNIQ times, (i.e. I=1,2,...,NUNIQ).
Examples of lines 3 and 4:
doublet OPENTYPE 1
1.0 1 1
triplet OPENTYPE 1
1.0 2 1 2
quartet OPENTYPE 1
1.0 3 1 2 3
open-shell singlet
OPENTYPE 2
1.0 1 1
1.0 1 2
5. FORMAT(A8)
NOSYM = blank symmetry will be used
= NOSYM symmetry turned off
________________________________________________________________________
1 10. DRT
---
A. Distinct Row Table program for the shape driven GUGA CI system.
This program will read a flexible input format of orbital codes,
rearrange the orbitals to a form suitable for the CI and generate
all the arrays needed to describe the CI calculation and the
integral storage.
The CI program can handle arbitrary reference sets, arbitrary
excitation levels, and reference sets such as triples in a selected
space and, say, singles outside that space. There is essentially no
limit to the total number of unpaired electrons, orbitals or
configurations except for the computer time available.
B. References:
J. Paldus, J. Chem. Phys. 61 (1974) 5321.
I. Shavitt, Int. J. Quantum Chem. Symp. 11 (1977) 131; 12 (1978) 5
Interacting configurations:
A. Bunge, J. Chem. Phys. 53 (1970) 20.
C. F. Bender and H. F. Schaefer, J. Chem. Phys. 55 (1971) 4798.
C. Files required: INPUT (# DRT ####)
Temporary files used: none
Files generated: CHECK
FILE6
FILE58
D. Input format:
# DRT ####
1. FORMAT(A78)
LABEL title for DRT output (for print out only)
2. FORMAT(8I5)
OPTION(1) printing option (powers of two)
= 1 ... print DRT
= 2 ... print external weight arrays
= 4 ... print integral pointer arrays
= 8
= 16
= 32
OPTION(2) excitation level for excitations into virtual
orbitals (default 2, i.e. CISD)
IMPORTANT: If this option is set > 2 then you need
to set I34X = 3 in the GUGACI input.
OPTION(3) excitation level for references in orbitals
flagged '%' (default 0)
OPTION(4) interacting configurations only (default is 'Yes'
for one reference and % orbitals, 'No' for multi-
reference)
= 1 ... turn off limitation to spin interacting
space (default for multi-reference).
= 2 ... limit valence references (%) to those of
the same symmetry as the one reference.
= 3 ... symmetry limit the valence references but
use full spin-space.
OPTION(5) integral block size desired, in hundreds
(default block size = 300000)
OPTION(6) = 0 ... (default)
= 1 ... use 4-external arrays in CI
OPTION(7) = 0 ... (default)
= n ... reassign output to this unit number
OPTION(8) = 0 ... (default)
= m ... set fermi-level to m
3. FORMAT(3I5)
NSYM number of symmetry classes
NBF number of basis functions
NREFS number of references
Note that TCSCF-CI is considered to be only one reference.
4. Occupation codes (free format)
(REPEAT COUNT) (KEY) CODE SYMMETRY
REPEAT COUNT (optional) is the number of identical orbitals
KEY (optional) is % for valence excitation orbitals
/ for orbitals differing in
different references
CODE is one of:
FZC frozen core
FZV frozen virtual
COR restricted core (integrals are transformed)
VIR restricted virtual
DOC doubly occupied
UOC virtual
ALP alpha occupancy (spin increase)
BET beta occupancy (spin decrease)
SPE special orbitals to be defined later in input
(used for open-shell singlets and TCSCF-CI)
SYMMETRY is a number (1-8) identifying the irreducible rep.
that the orbital belongs to.
N.B. It is mandatory to number in Cotton's way
(see the description for the SCF program).
For example: FZC1 2%DOC1 3 %UOC1 15UOC3 (blanks ignored)
*********************************************************************
* *
* For CI gradients you MUST use COR and VIR rather than FZC and *
* FZV. *
* For CC gradients you can only use DOC and UOC at present, and *
* the specification MUST be in C1 symmetry. *
* *
*********************************************************************
5. If NREFS > 1 :
Extra codes for references greater than the first.
Give only codes for those orbitals corresponding to those
flagged with a '/' in section 4. For example, if NREFS = 3,
three configurations from two orbitals could be as follows:
line 4. /DOC1 3DOC1 /UOC1 5UOC1 configuration 1
orbital A B
line 5. UOC1 DOC1 configuration 2
orbital A B
line 5. ALP1 BET1 configuration 3
orbital A B
6. If there are special codes, the program needs to know how to
handle them. The possibilities are:
GVB Placing this in the first three positions of this line
causes a two-reference interacting calculation to
be run. This should be used for closed shell TCSCF-CI
calculations.
OSS for open-shell singlet
MAT to enter matrix, etc. Additional lines required are |
7. FORMAT(3I5) |
number of electrons in special orbitals, spin*2, and |
total symmetry of special orbitals. |
8. FORMAT(4(4I1,1X)) |
matrix of excitations into orbitals, dimension |
4**number of special orbitals. Entered as a multi- |
dimensional array in Fortran, leftmost index giving |
case (1-4) for walk for first special orbital, etc. |
Example, for GVB pair: 0000 0111 0111 0112 |
________________________________________________________________________
1 11. TRANS
-----
A. TRANS carries out the transformation of integrals from the symmetry
adapted atomic orbital basis to the molecular orbital basis
for use in CI calculations.
B. References:
C. F. Bender J. Comput. Phys. 9 (1972) 547.
C. Files required: INPUT
FILE30
FILE34
FILE58
Temporary files used: FILE91
FILE93
FILE95
Files generated: CHECK
FILE6
FILE52 MO integrals DRT ordering
D. Input required: none
________________________________________________________________________
1 12. CISORT
------
A. CISORT sorts the integrals in the MO basis into the correct order
for the GUGA CI calculation.
B. References: none
C. Files required: INPUT
FILE52
FILE58
Temporary files used: FILE99
Files updated: FILE52
Files generated: CHECK
FILE6
D. Input required: none
________________________________________________________________________
1 13. GUGACI
------
A. GUGACI calculates the CI energy. The CI Hamiltonian matrix is
constructed using the shape driven graphical unitary group
approach for CISD. For higher excitations, the loop driven GUGA
approach is used. The Davidson correction is also calculated
(note, however, that the value printed out for the Davidson
correction is only correct for one reference CISD wavefunctions).
B. References:
GUGA-CI:
I. Shavitt, Int. J. Quantum Chem. Symp. 11 (1977) 131.
I. Shavitt, Int. J. Quantum Chem. Symp. 12 (1978) 5.
B. R. Brooks and H. F. Schaefer, J. Chem. Phys. 70 (1979) 5092.
P. Saxe, D. J. Fox, H. F. Schaefer and N. C. Handy, J. Chem.
Phys. 77 (1982) 5584.
Davidson correction:
E. R. Davidson, J. Comput. Phys. 17 (1975) 87.
S. R. Langhoff, E. R. Davidson, Int. J. Quantum Chem. 8 (1974) 61.
C. Files required: INPUT (# CI #####)
FILE52
FILE58
Temporary files used: FILE94
FILE99
Files generated: CHECK
FILE6
FILE54 CI coefficients
FILE95
D. Input format:
# CI ##### (if no input found uses defaults)
1. FORMAT(9I5)
(1) IGUESS = 0 ... (default) unit vector used to start the
CI iteration
= n ... read 'n' vectors from FILE54 to start.
(2) MXITER maximum iterations per root sought (10 with
2 roots gives 20 total)
Default 10 (15 recommended for CI gradients)
(3) IROOTI first root sought. Defaults to 1 (lowest)
without starting vectors for lower roots.
(4) NROOTS total number of roots to be sought
(default 1)
(5) NTOL = 0 ... convergence on CI vector = 10**-8
= n ... convergence on CI vector = 10**-n
(n=10 recommended for CI gradients)
(6) IRSTRT = -1 ... first iteration, save restart data on FILE95
= 0 ... no restart to be attempted (default)
= 1 ... attempt restart from prior run (requires
FILE95 from last run)
(7) I34X = 0 ... use 3 and 4 external vectorized routines
(default)
= 3 ... activates loop driven algorithm
Required for calculations above singles and
doubles
(8) ILVFRM = 0 ... (default)
= n ... value to set fermi level to
(9) IPRINT = 0 ... normal printing
= 1-2 .. more output
________________________________________________________________________
1 14. ONEPDM
------
A. ONEPDM may be used to construct the CI one particle density matrix,
to form the CI natural orbitals, and to perform an analysis of the
CI wavefunction.
B. References: none
C. Files required: INPUT (# ONEPDM #)
FILE30
FILE40 (if PRPFLG > 0)
FILE54
FILE58
Temporary files used: none
Files updated FILE30 (if PRPFLG = 2)
FILE40 (if PRPFLG = 1)
Files generated: CHECK
FILE6
FILE50 (if PRPFLG > 0)
FILE51 (called OPDM48) (if PRPFLG > 0)
D. Input format:
# ONEPDM # (default values used if # ONEPDM # is not found)
1. FORMAT(5I5,2X,A3)
PRINT = 0 ... no additional printing (default)
= 1 ... print the 1-PDM to FILE6 also
= 2 ... print the 1-PDM and NO-MO matrix to FILE6
= 3 ... print the 1-PDM, NO-MO matrix, and NO-SO
matrix to FILE6
MAX the MAX most important configurations are
displayed (default --- 20)
PRPFLG = -1 ... display the most important configurations
only (default)
= 0 ... in addition to the above, form the 1-PDM
and print the populations in the MO's
= 1 ... in addition to the above, diagonalize the
1-PDM to obtain the natural orbitals in
terms of the MO's, SO's, and AO's. The AO
1-PDM is then written to the master file
(FILE40).
= 2 ... in addition to the above, write the natural
orbitals (relative to the SO's) over the SCF
vector in FILE30
ROOTI the number of the first CI root for which
ONEPDM is to be run (default --- 1)
ROOTF the number of last CI root for which
ONEPDM is to be run (default --- 1)
PGROUP the point group of the molecule (e.g. C2V,
CS, etc.) for use in labelling MO's by the
correct irreducible representations
(default: D2H, C2V, CS, or C1).
________________________________________________________________________
1 15. TWOPDM
------
A. TWOPDM constructs the CI two particle density matrix for use in
calculating CI energy gradients.
B. References: see NGRCPHF
C. Files required: INPUT (# TWOPDM #)
FILE54
FILE58
Temporary files used: none
Files updated: FILE54
Files generated: CHECK
FILE6
FILE53
D. Input format:
# TWOPDM # (default values used if # TWOPDM # not used)
1. FORMAT(2I5)
IGUESS root of CI used to calculate 2-PDM
(default = 1).
IPRINT = 0 ... minimum printing
> 0 ... more printing
________________________________________________________________________
1 16. LAGTR
-----
A. LAGTR constructs the Lagrangian matrix and performs the first part
of the back transformation of the 2-PDM for determining correlated
energy gradients.
B. References: see NGRCPHF
C. Files required: INPUT (# LAGTR ##)
FILE30
FILE40
FILE52
FILE53
FILE54
FILE58
Temporary files used: FILE91
FILE93
FILE95
Files updated: FILE40 (if MASTER = 1)
Files generated: CHECK
FILE6
FILE47
FILE55
FILE71
FILE85 (for MONGO) |
D. Input format:
# LAGTR ##
1. FORMAT(3I5)
MASTER = 0 ... default
= 1 ... write the Lagrangian in DRT ordering to
FILE40
IDFILE not used
ICCSD = 0 ... CI gradient calculation (default)
= 1 ... CCSD gradient calculation
________________________________________________________________________
1 17. DERTFK
------
A. DERTFK completes the back transformation of the 2-PDM and calculates
the derivative AO integrals for correlated wavefunctions.
B. References: see NGRCPHF
C. Files required: INPUT (# NEWDER #)
FILE30
FILE49
FILE55
Temporary files used: none
Files updated: FILE30
Files generated: CHECK
FILE6
FILE42
FILE78 (if INTOUT = 1)
FILE79 (if INTOUT = 1)
D. Input format:
# NEWDER #
1. FORMAT(A5,5X,A6,I5)
CALTYP = CI for CI and CC gradients
= GVBCI for TCSCF-CI
and excited state SCF-CI gradients |
LEVEL = FIRST first derivatives
= SECOND * not available at present
IPRINT = 0 minimum printing
= 2 more output
= 4
= 8
= higher powers of 2
2. FORMAT(A8,2I5)
NOSYM = blank symmetry will be used
= NOSYM symmetry turned off
INTOUT = 0 normal run
= 1 write out derivative one-electron integrals
to FILE79 and derivative two-electron
integrals to FILE78
INFOUT (not used)
________________________________________________________________________
1 18. NGRCPHF
-------
A. NGRCPHF solves the coupled-perturbed Hartree-Fock equations for
correlated wavefunctions and completes the calculation of
the gradient and the dipole moment.
B. References:
B. R. Brooks, W. D. Laidig, P. Saxe, J. D. Goddard, Y. Yamaguchi
and H. F. Schaefer, J. Chem. Phys. 72 (1980) 4652.
Y. Osamura, Y. Yamaguchi and H. F. Schaefer, J. Chem. Phys. 77
(1982) 383.
N. C. Handy and H. F. Schaefer, J. Chem. Phys. 81 (1984) 5031.
J. E. Rice, R. D. Amos, N. C. Handy, T. J. Lee and H. F. Schaefer,
J. Chem. Phys. 85 (1986) 963.
Y. Osamura, Y. Yamaguchi and H. F. Schaefer, Theor. Chim. Acta 72
(1987) 71.
C. Files required: INPUT (# GRCPHF #)
FILE30
FILE42
FILE47
FILE52
FILE54
FILE58
FILE59
FILE69
Temporary files used: FILE94
FILE96
FILE98
Files updated: FILE30
Files generated: CHECK
FILE6
FILE11
FILE86 (for MONGO) |
FILE87 (for MONGO) |
FILE88 (for MONGO) |
D. Input format:
# GRCPHF # (default values used if no # GRCPHF # found)
1. FORMAT(3I5)
IPRINT = 0 ... minimum printing (default)
= 2 ... more output
= 4 (or higher powers of 2)
ICIDIP not used (CPHF correction to dipole moment
will always be calculated)
Note: this means that for the current version of
NGRCPHF to work, the program CIPROP must be run
before NGRCPHF.
ICCSD = 0 ... CI gradient calculation (default)
= 1 ... CCSD gradient calculation
________________________________________________________________________
1 19. CICPHF
------
A. CICPHF should be used in place of NGRCPHF for calculating TCSCF-CI
gradients.
... and excited state SCF-CI gradients. |
B. References:
T. J. Lee, W. D. Allen and H. F. Schaefer, J. Chem. Phys. 87
(1987) 7062.
W. D. Allen and H. F. Schaefer, J. Chem. Phys. 87 (1987) 7076.
C. Files required: INPUT (# GRCPHF #)
FILE30
FILE42
FILE47
FILE54
FILE58
FILE59
Temporary files used: FILE94
FILE96
FILE98
Files updated: FILE30
Files generated: CHECK
FILE6
FILE11
D. Input format:
# GRCPHF #
1. FORMAT(3I5)
IPRINT = 0 ... minimum printing (default)
= 2 ... more output
= 4 (or higher powers of 2)
ICIDIP = 0 ... no dipole moment calculated
= 1 ... CPHF correction to dipole moment calculated
KSPE = 0 ... normal run (TCSCF-CI)
or excited state SCF-CI) |
= 2 ... two special orbitals |
This option is to allow a three reference CI |
gradient to be calculated (NREFS = 3 in the |
DRT input). |
|
2. If KSPE = 2: |
FORMAT(2I5) |
LSPE(1), LSPE(2) integer labels of the special orbitals |
(i.e. the numbers assigned to them |
as in the DRT input) |
The three references are: |
(SPE1)2, (SPE2)2, (SPE1)(SPE2) (open-shell singlet) |
________________________________________________________________________
1 20. CCTRANS
-------
A. CCTRANS transforms integrals from the AO basis (FILE34) to the MO
basis (FILE78) using an intermediate file (FILE77).
B. References: none
C. Files required: INPUT
FILE30
FILE34
Temporary files used: FILE77
Files generated: CHECK
FILE6
FILE67
FILE78
D. Input required: none
________________________________________________________________________
1 21. NCCSRT
------
A. NCCSRT sorts the integrals in FILE78 into different groups
(FILE60-66) as used by NCC9.
B. References: none
C. Files required: INPUT
FILE30
FILE78
Temporary files used: none
Files generated: CHECK
FILE6
FILE60
FILE61
FILE62
FILE63
FILE64
FILE65
FILE66
D. Input required: none
________________________________________________________________________
1 22. NCC9
------
A. NCC9 calculates the closed shell coupled-cluster energy.
CCSD is (2 to 8 times) more expensive than CISD. The gradient time
overhead is about 100% of the energy time.
CCSDT-1 calculations are expensive. They scale as N (the number of
basis functions) to the seventh power, one order of magnitude bigger
than CCSD. The gradient time overhead is about 200%.
Savings in CC are non-linear with respect to the number of
irreducible representations, i.e. the higher your symmetry point
group, the more you save.
B. References:
G. E. Scuseria, T. J. Lee and H. F. Schaefer, Chem. Phys. Lett.
130 (1986) 236.
G. E. Scuseria and H. F. Schaefer, Chem. Phys. Lett. 142 (1987)
354.
G. E. Scuseria, A. C. Scheiner, T. J. Lee, J. E. Rice and H. F.
Schaefer, J. Chem. Phys. 86 (1987) 2881.
G. E. Scuseria, C. L. Janssen and H. F. Schaefer, J. Chem. Phys.
89 (1988) 7382.
C. Files required: INPUT (# CCSD ###)
FILE30
FILE60
FILE61
FILE62
FILE63
FILE64
FILE65
FILE66
FILE67
Temporary files used: FILE97
FILE98
FILE99
Files generated: CHECK
FILE6
FILE68
FILE69 (CC vector)
FILE81
D. Input format:
# CCSD ###
1. FORMAT(A80)
TITLE Let your imagination fly.
2. FORMAT(7I5)
DIIS1 = 0 ... (keep zero)
DIIS2 = 0 ... (keep zero)
DIIS3 = 0 ... (keep zero)
DIIS4 = 0 ... (keep zero)
DIIS5 = 0 ... (keep zero)
FLDIIS = 0 ... normal use of DIIS
= 2 ... turn DIIS off
IRSTR = 0 ... normal run
= 1 ... restart job (needs FILE69)
3. FORMAT(2I5)
CONVI = 0 ... convergence = 10**-7
= n ... convergence = 10**-n
MAXIT = 0 ... max number of iterations = 20
= n ... max number of iterations = n (50 is enough)
4. FORMAT(I2,1X,I2)
CORS the number of COR orbitals (see definition in DRT
description) in each irreducible representation
VIRS the number of VIR orbitals (see definition in DRT
description) in each irreducible representation
No DOCs or UOCs are needed.
For a CCSD gradient calculation, no frozen orbitals are
allowed at present, i.e. CORS and VIRS must be equal to zero.
.....repeat this line for each irreducible representation
5. FORMAT(A4)
OPTION = CCSD coupled cluster single and double excitations
= SDT1 coupled cluster single, double and linearized
triple excitations
= CHEK ask Guscus |
= MP2 " " |
= CCD " " |
= LCCD " " |
= LCSD " " |
= VAR2 " " |
________________________________________________________________________
1 23. NZCCSD
------
A. NZCCSD solves the coupled-perturbed coupled cluster equations for
CCSD and CCSDT-1 wavefunctions.
B. References:
A. C. Scheiner, G. E. Scuseria, T. J. Lee, J. E. Rice and H. F.
Schaefer, J. Chem. Phys. 87 (1987) 5361.
G. E. Scuseria and H. F. Schaefer, Chem. Phys. Lett. 146 (1988)
23.
C. Files required: INPUT (# ZCCSD ##)
FILE30
FILE60
FILE61
FILE62
FILE63
FILE64
FILE65
FILE66
FILE68
Temporary files used: FILE91
FILE92
FILE93
FILE94
FILE95
FILE96
FILE97
FILE98
FILE99
Files updated: FILE69
Files generated: CHECK
FILE6
FILE82
D. Input format:
# ZCCSD ##
1. FORMAT(A80)
TITLE
2. FORMAT(2I5)
CONVI = 0 ... convergence = 10**-10
= n ... convergence = 10**-n
MAXIT = 0 ... max number of iterations = 30
= n ... max number of iterations = n (30 is enough)
3. FORMAT(A4)
OPTION = NORM normal run
= RSTR restart (needs FILE69)
________________________________________________________________________
1 24. CCDMAT3
-------
A. CCDMAT3 calculates the effective one- and two-particle density
matrices for CCSD and CCSDT-1 wavefunctions.
B. References: see NZCCSD
C. Files required: INPUT
FILE30
FILE69
FILE81
FILE82
Temporary files used: none
Files generated: CHECK
FILE6
FILE68
D. Input required: none
________________________________________________________________________
1 25. CCTODRT
-------
A. CCTODRT sorts the integrals and density matrices from CC to DRT
ordering.
B. References: see NZCCSD
C. Files required: INPUT
FILE30
FILE58
FILE68
FILE78
Temporary files used: FILE91
Files generated: CHECK
FILE6
FILE52
FILE53
D. Input format:
No input is required, but the input for the DRT program must be in
C1 symmetry and consist only of DOCs and UOCs.
________________________________________________________________________
1 26. MASTER
------
A. MASTER uses the SCF information to form the master file
(FILE40) for use in subsequent programs.
The master file contains the necessary information (e.g. SCF
eigenvectors, sorted eigenvectors, parameters, constants, etc.)
to calculate SCF analytical derivatives and properties.
B. References: none
C. Files required: INPUT (# MASTER #)
FILE30
FILE34
Temporary files used: none
Files generated: CHECK
FILE6
FILE36
FILE40
D. Input format: (almost the same as for DERIV)
# MASTER #
1. FORMAT(3(A8,2X))
SCFTYP = CLSCF ... for closed shell SCF
= GRSCF ... for CI gradient or open shell SCF
Always use GRSCF for CI and CC gradients
even if a closed shell SCF reference
wavefunction is employed.
= TCSCF ... for TCSCF
... and excited state |
= MCSCF * not available at present
CITYP = SCF for SCF derivatives
= CI for CI and CC derivatives
= MCSCF * not available at present
DERTYP order of derivative
= FIRST
= SECOND
= THIRD
2. FORMAT(I5)
IPRINT = 0 ... normal printing
= 1-6 .. more output
3. If SCFTYP = GRSCF and IOPEN (in # SCF ####) = 1:
FORMAT(A8,2X,I5)
OPTYPE = OPENTYPE
NUNIQ = 1 ... for doublet, triplet or any high-spin open
shell
= 2 ... for open-shell singlet
Lines 3 and 4 are NOT needed for TCSCF
4. If SCFTYP = GRSCF and IOPEN (in # SCF ####) = 1:
FORMAT(F10.5,10I5)
GOCC(I) occupation in Ith shell (type) - usually 1.0
(i.e. one electron in each orbital)
LL number of open-shells in Ith shell (type)
MOPN(I,J),J=1,LL list of numbers of open-shells in Ith shell
(type).
Usually MOPN(I,J)=1,2,...,LL
.....repeat this line NUNIQ times, (i.e. I=1,2,...,NUNIQ).
Examples of lines 3 and 4:
doublet OPENTYPE 1
1.0 1 1
triplet OPENTYPE 1
1.0 2 1 2
quartet OPENTYPE 1
1.0 3 1 2 3
open-shell singlet
OPENTYPE 2
1.0 1 1
1.0 1 2
5. FORMAT(A8)
TAPE = blank form FILE36
= NOFILE36 do not form FILE36
________________________________________________________________________
1 27. MAKE37
------
A. MAKE37 reads the integral file (FILE34) and forms a supermatrix
PK-integral file (FILE37).
B. References:
PK-file method:
R. C. Raffenetti, Chem. Phys. Lett. 20 (1973) 335.
C. Files required: INPUT (# MAKE37 #)
FILE34
Temporary files used: none
Files generated: CHECK
FILE6
FILE37
D. Input format:
# MAKE37 # (if not found, program uses defaults)
1. FORMAT(3I5)
ITOLER = 0 ... cutoff for integrals = 10**-12
= n ... cutoff for integrals = 10**-n
ITEST = 0 ... normal
= 1 ... writes supermatrix to output file
IPRNT = 0 ... normal printing
= 1-5 .. more output
________________________________________________________________________
1 28. CPCLAO / CPCLAOS
----------------
A. CPCLAO solves the coupled perturbed Hartree-Fock equations for
closed shell SCF wavefunctions in the AO basis.
CPCLAOS is the same as CPCLAO, except that it uses the super matrix
(PK-file) formulation.
B. References:
J. Gerratt and I. M. Mills, J. Chem. Phys. 49 (1968) 1719.
J. A. Pople, R. Krishnan, H. B. Schlegel and J. S. Binkley, Int.
J. Quant. Chem. Symp. S13 (1979) 225.
AO basis:
Y. Osamura, Y. Yamaguchi, P. Saxe, D. J. Fox, M. A. Vincent and
H. F. Schaefer, J. Mol. Struct. 103 (1983) 183.
C. Files required: INPUT (# CPHFAO #)
FILE36
FILE37
FILE40
FILE42
FILE43
Temporary files used: FILE91
Files generated: CHECK
FILE6
FILE11
FILE15
FILE17
FILE44
D. Input format:
# CPHFAO #
1. FORMAT(7I5)
ITEST = 0 ... no test |
= 1 ... test MO integrals and derivative integrals |
IORB = 0 ... no orbital energy derivatives |
= 1 ... calculate orbital energy derivatives |
IPOL = 0 ... no dipole derivatives or polarizabilities
(uses a little less memory)
= 1 ... calculate dipole derivatives and
polarizabilities
ICONV = 0 ... convergence on the CPHF equations = 10**-10
= n ... convergence on the CPHF equations = 10**-n
ICORE not used for closed shell |
IPRNT = 0 ... minimum printing
= 1-6... more output
IHYPR = 0 ... no hyperpolarizabilities
= 1 ... calculate hyperpolrizabilities
________________________________________________________________________
1 29. CPGRAO / CPGRAOS
----------------
A. CPGRAO solves the coupled perturbed Hartree-Fock equations for
general open-shell SCF wavefunctions in the AO basis.
CPGRAOS is the same as CPGRAO, except that it uses the super matrix
(PK-file) formulation.
B. References:
J. Gerratt and I. M. Mills, J. Chem. Phys. 49 (1968) 1719.
AO basis:
Y. Osamura, Y. Yamaguchi, P. Saxe, D. J. Fox, M. A. Vincent and
H. F. Schaefer, J. Mol. Struct. 103 (1983) 183.
C. Files required: INPUT (# CPHFAO #)
FILE36
FILE37
FILE40
FILE42
FILE43
Temporary files used: FILE91
Files generated: CHECK
FILE6
FILE11
FILE15
FILE17
FILE44
D. Input format:
# CPHFAO #
1. FORMAT(6I5)
ITEST = 0 ... no test |
= 1 ... test MO integrals and derivative integrals |
IORB = 0 ... no orbital energy derivatives |
= 1 ... calculate orbital energy derivatives |
IPOL = 0 ... no dipole derivatives or polarizabilities
(uses a little less memory)
= 1 ... calculate dipole derivatives and
polarizabilities
ICONV = 0 ... convergence on the CPHF equations = 10**-10
= n ... convergence on the CPHF equations = 10**-n
ICORE = 0 ... no calculations for core-core pairs |
1 ... calculate core-core pairs |
IPRNT = 0 ... minimum printing
= 1-6... more output
________________________________________________________________________
1 30. CPTCAO / CPTCAOS
----------------
A. CPTCAO solves the coupled perturbed Hartree-Fock equations for
closed shell TCSCF wavefunctions in the AO basis.
... and for excited state SCF wavefunctions. |
CPTCAOS is the same as CPTCAO, except that it uses the super matrix
(PK-file) formulation.
B. References: see CPGRAO
Y. Yamaguchi, Y. Osamura and H. F. Schaefer, J. Am. Chem. Soc.
105 (1983) 7507.
C. Files required: INPUT (# CPHFAO #)
FILE36
FILE37
FILE40
FILE42
FILE43
Temporary files used: FILE91
FILE92
Files generated: CHECK
FILE6
FILE11
FILE15
FILE17
FILE44
D. Input format:
# CPHFAO #
1. FORMAT(6I5)
ITEST = 0 ... no test |
= 1 ... test MO integrals and derivative integrals |
IORB = 0 ... no orbital energy derivatives |
= 1 ... calculate orbital energy derivatives |
IPOL = 0 ... no dipole derivatives or polarizabilities
(uses a little less memory)
= 1 ... calculate dipole derivatives and
polarizabilities
ICONV = 0 ... convergence on the CPHF equations = 10**-10
= n ... convergence on the CPHF equations = 10**-n
ICORE = 0 ... no calculations for core-core pairs |
1 ... calculate core-core pairs |
IPRNT = 0 ... minimum printing
= 1-6... more output
________________________________________________________________________
1 31. DER3RD
------
A. DER3RD calculates the third derivative integrals.
At present, this program is restricted to a maximum of seven atoms.
B. References:
J. F. Gaw, Y. Yamaguchi, H. F. Schaefer and N. C. Handy, J. Chem.
Phys. 85 (1986) 5132.
J. F. Gaw, Y. Yamaguchi, R. B. Remington, Y. Osamura and H. F.
Schaefer, Chem. Phys. 109 (1986) 237.
C. Files required: INPUT (# DER3RD #)
FILE30
FILE40
Temporary files used: FILE91
FILE99
Files generated: CHECK
FILE6
FILE38
FILE42
FILE45
FILE46
D. Input format:
# DER3RD #
1. FORMAT(3(A8,2X))
SCFTYP = CLSCF
= GRSCF
= TCSCF * not available at present
= MCSCF * not available at present
CITYP = SCF
DERTYP = THIRD
2. FORMAT(3I5)
IPRNT = 0 ... normal printing
= 1-3 .. more output
IDRVT = 0 or 1 store only the two electron first derivative
integrals on FILE38
= 2 ... store the two electron first and second
derivative integrals on FILE38
= 3 ... store the two electron first, second and
third derivative integrals on FILE38
IDRVF = 0 or 2 calculate first and second derivative fock
matrices
= 1 ... calculate only first derivative fock matrices
= 3 ... calculate first, second and third derivative
fock matrices
________________________________________________________________________
1 32. CL3RD
-----
A. CL3RD completes the calculation of the third derivatives in the
AO basis for closed shell SCF wavefunctions.
CL3RD should only be used for non-degenerate closed shell systems.
B. References:
J. F. Gaw, Y. Yamaguchi, H. F. Schaefer and N. C. Handy, J. Chem.
Phys. 85 (1986) 5132.
C. Files required: INPUT (# SCF3RD #)
FILE38
FILE40
FILE44
FILE45
FILE46
Temporary files used: FILE91
Files generated: CHECK
FILE6
FILE20
FILE21 (if ITEST = 1) |
FILE22 (if ITEST = 1) |
D. Input format:
# SCF3RD #
1. FORMAT(4I5)
ITEST = 0 ... normal run |
= 1 ... store F3A matrix on FILE21 and F3M matrix on |
FILE22 |
IORB = 0 ... not used at present |
IPOL = 0 ... should match CPHFAO input
IPRNT = 0 ... normal printing
= 1-6 .. more output
________________________________________________________________________
1 33. GR3RD
-----
A. GR3RD completes the calculation of the third derivatives in the
AO basis for general open shell SCF wavefunctions.
GR3RD should also be used for degenerate closed shell systems.
B. References:
J. F. Gaw, Y. Yamaguchi, R. B. Remington, Y. Osamura and H. F.
Schaefer, Chem. Phys. 109 (1986) 237.
C. Files required: INPUT (# SCF3RD #)
FILE36
FILE38
FILE40
FILE42
FILE44
FILE45
FILE46
Temporary files used: FILE91
FILE92
Files generated: CHECK
FILE6
FILE20
FILE21 (if ITEST = 1) |
FILE22 (if ITEST = 1) |
D. Input format:
# SCF3RD #
1. FORMAT(4I5)
ITEST = 0 ... normal run |
= 1 ... store F3A matrix on FILE21 and F3M matrix on |
FILE22 |
IORB = 0 ... not used at present |
IPOL = 0 ... should match CPHFAO input
IPRNT = 0 ... normal printing
= 1-6 .. more output
________________________________________________________________________
1 34. GVBSCF
------
A. GVBSCF calculates SCF energies for the generalized valence bond
(GVB) and paired-excited multi-configuration SCF (PE MCSCF) wave-
functions.
B. References:
F. M. Bobrowicz and W. A. Goddard, in "Methods of Electronic
Structure Theory", ed. H. F. Schaefer (Plenum, New York, 1977)
R. Carbo and J. M. Riera, "A General SCF Theory" (Topics in Current
chemistry, Springer, Berlin, 1978)
C. Files required: INPUT (# SCF ####, # PEX ####)
(# EFIELD #, # SUPMX ##, |
# MOFLIP #, # COUPLING #) |
FIELD
FILE30
FILE34
Temporary files used: FILE92
Files updated: FILE30 MO coefficients
Files generated: CHECK
FILE6
FILE16
FILE36
D. Input format:
# SCF ####
1. FORMAT(A40)
ETIQ title for SCF output (free field)
(for print out only)
2. FORMAT(10I5)
(1) NLVS......... number of iterations that the level shift is
to be used
(2) ICONV = 0 ... convergence on density matrix = 10**(-8)
(default)
(n=9 recommended for single point energies,
n=10 to 12 recommended for derivatives)
= n ... convergence on density matrix = 10**(-n)
(3) IWF = 0 ... closed shell
1 ... open shell
2 ... closed shell TCSCF (GVB-1)
3 ... open shell TCSCF (GVB-1)
4 ... excited same symm open-shell singlet
5 ... closed-shell GVB
6 ... open-shell GVB
10 ... complete PEMCSCF
11 ... selected PEMCSCF (configurations read
from # PEX ####)
12 ... paired-excited MCSCF, only double excitations
... Note: if IWF<0 (above list is still valid for its absolute
value), a pseudoeigenvalue method is used that works
fine for convergence <10**-4. However, it does not
carry convergence further than 10**-10.
(4) IREAD = 0 ... core guess
1 ... guess in FILE30
(5) IROOT........ root desired in a CASSCF. default 2
(6) IORDER = 0 ... no reordering
1 ... reorder according to integer vector input
(see below)
(7) IPRNT = 0 ... normal printing
= -1 ... less printing
= 1-7... more printing
(8) NODIIS....... 10**(-NODIIS) is the error value at which DIIS
starts
(9) NDIIS........ number of iterations to extrapolate Fock
matrices if DIIS wants to be performed. NDIIS
must be positive for closed-shell calculations
and negative for open-shell and MCSCF calcula-
tions. If pseudoeigenvalues methods are used
(IOPEN<0), NDIIS must be positive.
(10) NITER........ maximum number of SCF iterations (default: 40)
3. If IORDER = 1
FORMAT(14I5)
IREO(I),I=1,NBASIS
...... List the new order of MO's
Number orbitals as if in C1 symmetry always
4. FORMAT(8I5)
IDOCC(I),ISOCC(I), K1, K2, K3, K4, K5
..... repeat NST times (i.e. I=1, NST)
NST is the number of irreducible representations
5. FORMAT(8I5)
YCL(I), YOP(I), (YGVB(I,J),J=1,3), YPEX(I), YELPEX(I)
..... repeat NST times (i.e. I=1, NST)
6. FORMAT(2F20.10)
ALPHA(I) open shell coupling coefficient (alpha)
BETA(I) open shell coupling coefficient (beta)
.....repeat this line MM*(MM+1)/2 times, where MM is the number of
symmetry irreducible representations containing singly-
occupied MO's
Examples:
for doublet:
0.0 -1.0
for triplet:
0.0 -1.0
0.0 -1.0
0.0 -1.0
for open-shell singlet:
0.0 -1.0
0.0 3.0
0.0 -1.0
for TCSCF: (constants supplied by program for TCSCF but a
dummy set still needed)
0.0 0.0
0.0 -1.0
0.0 0.0
for C1 symmetry, only one set of coupling coefficients is
possible (i.e. it is not possible to do open-shell
singlets or TCSCF in C1 symmetry with this program):
0.0 -1.0
for high-spin open-shell wavefunctions, the values of alpha
and beta are always 0.0 and -1.0, respectively.
7. FORMAT(2F20.10)
XDUM ....... damping factor (not in use)
ZLVS ....... level shift
If IWF = 11:
# PEX ####
1. FORMAT(*)
NCPEX ........ number of configurations to be entered
2. FORMAT(20I1)
INDPEX(IU,I),IU=1,NORPEX
.... repeat NCPEX times (i.e. I = 1, NCPEX)
.... one configuration per line.
.... a 1 means occupied orbital, a 0 vacant.
# EFIELD # |
1. FORMAT(*) |
IFIELD = 0 ... no electric field |
= 1 ... electric field to be included |
DISPLA ......... the displacement to be used (? not used?) |
.... electric field effect |
|
|
FIELD (logical unit 14) |
1. IF IFIELD .NE. 0 |
FORMAT(*) |
FX |
FY |
FZ |
..... FX, FY, and FZ are displacement |
|
|
|
|
# SUPMX ## |
1. FORMAT(2I5) |
ISUPMX = 0 ... no supermatrix |
= 1 ... the two-electron integrals are set up |
in supermatrix form |
INOSYM = 1 ... must equal 1 to run third-derivative |
program PX3RD |
|
|
|
# MOFLIP # |
1. FORMAT(*) |
IIIJ |
JJJI |
..... switching of a pair of MO's (IIIJ and JJJI) if necessary |
after diagonalization by COUPOP. This is necessary if two |
or more open shells are run in the same symmetry, and the |
ordering of them is inverse to their occupations. |
|
|
# COUPLING # |
1. FORMAT(*) |
OCC(I),I=1,NSHL ... |
ALFA(I,J),J=1,I),I=1,NSHL |
BETA(I,J),J=1,I),I=1,NSHL |
.... enter coupling parameters: F(I),ALPHA(I,J),BETA(I,J) (Goddard's |
way) in free format, lower triangular form. |
________________________________________________________________________
1 35. GVBDER / GVBDER2
----------------
A. GVBDER calculates first derivatives for GVB and paired-excited
MCSCF wavefunctions
GVBDER2 calculates second derivatives for GVB and paired-excited
MCSCF wavefunctions
B. References:
C. Files required: INPUT (# DERIV ##)
FILE30
Temporary files used: none
Files generated: CHECK
FILE6
FILE11
D. Input format:
# DERIV ##
1. FORMAT(3(A8,2X))
SCFTYP = TCSCF ... for TCSCF
= GVBSCF ... for GVBSCF
CITYP = SCF ... for SCF derivatives
DERTYP = FIRST ... (for program GVBDER)
= SECOND ... (for program GVBDER2)
2. FORMAT(I5)
IPRINT = 0 ... normal printing
= 1-6 .. more output
3. FORMAT(A8)
NOSYM = blank symmetry will be used
= NOSYM symmetry turned off
_______________________________________________________________________
1 36. MASTERPX
--------
A. MASTERPX uses Pitzer's SCF information to form the master file
(FILE40) for use in subsequent programs.
This version is specifically for PEMCSCF calculations.
B. References: none
C. Files required: INPUT
FILE30
FILE34
Temporary files used: none
Files generated: CHECK
FILE6
FILE36
FILE40
D. Input format:
# MASTER #
1. FORMAT(3(A8,2X))
SCFTYP = TCSCF ... for TCSCF
= GVBSCF ... for GVBSCF
CITYP = SCF ... for SCF derivatives
DERTYP = FIRST ... (for program GVBDER)
= SECOND ... (for program GVBDER2)
2. FORMAT(I5)
IPRINT = 0 ... normal printing
= 1-6 .. more output
3. FORMAT(A8)
TAPE = blank form FILE36
= NOFILE36 do not form FILE36
________________________________________________________________________
1 37. FORM37
------
A. FORM37 reads the integral file (FILE34) and forms a supermatrix
integral file (FILE37) for PEMSCF second and third derivatives.
This version does not take into account any symmetry, but
eliminates P and Q when they are smaller than a threshold.
The maximum number of basis functions is 125.
B. References:
C. Files required: INPUT
FILE34
Temporary files used: none
Files generated: CHECK
FILE6
FILE37
D. Input required: none
________________________________________________________________________
1 38. CPPXAO
------
A. CPPXAO solves the coupled perturbed Hartree-Fock equations for
PEMCSCF wavefunctions in the AO basis.
It is similar to CPCLAO, but specifically for the GVB and the
paired-excited MCSCF wavefunctions.
B. References for PEMCSCF:
M. Duran, Y. Yamaguchi, R. B. Remington, and H. F. Schaefer,
Chem. Phys. 122 (1988) 201.
C. Files required: INPUT
FILE36
FILE37
FILE40
FILE42
FILE43
Temporary files used: FILE91
FILE92
Files generated: CHECK
FILE6
FILE15
FILE17
FILE44
D. Input format:
# CPHFAO #
1. FORMAT(6I5)
ITEST = 0 ... no test |
= 1 ... test MO integrals and derivative integrals |
IORB = 0 ... no orbital energy derivatives |
= 1 ... calculate orbital energy derivatives |
IPOL = 0 ... calculate dipole derivatives and
polarizabilities
= 1 ... no dipole derivatives or polarizabilities
(uses a little less memory)
ICONV = 0 ... convergence on the CPHF equations = 10**-10
= n ... convergence on the CPHF equations = 10**-n
ICORE not used for closed shell |
IPRNT = 0 ... minimum printing
= 1-6... more output
2. FORMAT(A8) |
RESTART = 'RESTART ' |
= blank |
________________________________________________________________________
1 39. CPTCAOX
-------
A. CPTCAOX solves the coupled perturbed Hartree-Fock equations for
TCSCF wavefunctions in the AO basis.
It is similar to CPTCAO, but specifically for the wave-
functions using the PEMCSCF formulation.
B. References for TCSCF:
Y.Yamaguchi, Y.Osamura, and H.F.Schaefer, J. Am. Chem. Soc. 105,
(1983) 7506.
M. Duran, Y. Yamaguchi, Y. Osamura, and H. F. Schaefer, J. Mol.
Struct. (Theochem), 163 (1988) 389.
C. Files required: INPUT
FILE36
FILE37
FILE40
FILE42
FILE43
Temporary files used: FILE91
FILE92
Files generated: CHECK
FILE6
FILE15
FILE17
FILE44
D. Input format:
# CPHFAO #
1. FORMAT(6I5)
ITEST = 0 ... no test |
= 1 ... test MO integrals and derivative integrals |
IORB = 0 ... no orbital energy derivatives |
= 1 ... calculate orbital energy derivatives |
IPOL = 0 ... calculate dipole derivatives and
polarizabilities
= 1 ... no dipole derivatives or polarizabilities
(uses a little less memory)
ICONV = 0 ... convergence on the CPHF equations = 10**-10
= n ... convergence on the CPHF equations = 10**-n
ICORE not used for closed shell |
IPRNT = 0 ... minimum printing
= 1-6... more output
2. FORMAT(A8) |
RESTART = 'RESTART ' |
= blank |
________________________________________________________________________
1 40. NEW3RD
------
A. NEW3RD calculates the third derivative integrals for GVB and
paired excited MCSCF wavefunctions.
At present, this program is restricted to a maximum of 7 atoms.
B. References:
J. F. Gaw, Y. Yamaguchi, H. F. Schaefer and N. C. Handy, J. Chem.
Phys. 85 (1986) 5132.
J. F. Gaw, Y. Yamaguchi, R. B. Remington, Y. Osamura and H. F.
Schaefer, Chem. Phys. 109 (1986) 237.
M. Duran, Y. Yamaguchi, Y. Osamura, and H. F. Schaefer, J. Mol.
Struct. (Theochem), 163 (1988) 389.
C. Files required: INPUT
FILE30
FILE40
Temporary files used: FILE91
Files generated: CHECK
FILE6
FILE11
FILE38
FILE42
FILE45
FILE46
D. Input format:
# NEW3RD #
1. FORMAT(3(A8,2X))
SCFTYP = TCSCF
= GVBSCF
CITYP = SCF
DERTYP = THIRD
2. FORMAT(3I5)
IPRNT = 0 ... normal printing
= 1-3 .. more output
IDRVT = 0 or 1 store only the two electron first derivatives
on FILE38
= 2 ... store the two electron first and second
derivatives on FILE38
= 3 ... store the two electron first, second and
third derivatives on FILE38
IDRVF = 0 or 2 calculate first and second derivative fock
matrices
= 1 ... calculate only first derivative fock matrices
= 3 ... calculate first, second and third derivative
fock matrices
________________________________________________________________________
1 41. PX3RD
-----
A. Analytical third derivative program for GVB-SCF wavefunctions.
B. References: (See NEW3RD).
C. Files required: INPUT
FILE30
FILE37
FILE38
FILE40
FILE42
FILE44 for first derivatives
FILE45
FILE46
FILE47 for bare lagrangians
FILE48 for second derivatives
Temporary files used: FILE91
FILE92
Files updated: FILE48
Files generated: CHECK
FILE6
FILE20
FILE21 |
FILE22 |
D. Input format:
# SCF3RD #
1. FORMAT(4I5)
ITEST = 0 ... normal run |
= 1 ... store F3A matrix on FILE21 and F3M matrix on |
FILE22 |
IORB = 0 ... not used at present |
IPOL = 0 ... should match CPHFAO input
IPRNT = 0 ... normal printing
= 1-6 .. more output
________________________________________________________________________
1 42. TC3RD
-----
A. Analytical third derivative program for TCSCF wavefunctions.
B. References: (See NEW3RD)
C. Files required: INPUT
FILE30
FILE37
FILE38
FILE40
FILE42
FILE44 for first derivatives
FILE45
FILE46
FILE47 for bare lagrangians
FILE48 for second derivatives
Temporary files used: FILE91
FILE92
Files updated: FILE48
Files generated: CHECK
FILE6
FILE20
FILE21 |
FILE22 |
D. Input format:
# SCF3RD #
1. FORMAT(4I5)
ITEST = 0 ... normal run |
= 1 ... store F3A matrix on FILE21 and F3M matrix on |
FILE22 |
IORB = 0 ... not used at present |
IPOL = 0 ... should match CPHFAO input
IPRNT = 0 ... normal printing
= 1-6 .. more output
________________________________________________________________________
1 43. DIPDERPX
--------
A. DIPDERPX calculates derivatives of SCF dipole moments.
This version is specifically for PEMCSCF wavefunctions (the
input format is exactly the same as for DIPDER)
B. References:
Y. Yamaguchi, M. Frisch, J. Gaw, H. F. Schaefer and J. S.
Binkley, J. Chem. Phys. 84 (1986) 2262.
C. Files required: INPUT
FILE30
Temporary files used: none
Files generated: CHECK
FILE6
FILE43
D. Input format:
# DIPDER #
1. FORMAT(I5)
IPRNT = 0 ... minimum printing
= 1-6 .. more output
________________________________________________________________________
1 44. INTCOS
------
A. INTCOS transforms geometries and gradients from Cartesian
coordinates to internal coordinates.
B. References: none
C. Files required: INPUT (# INTCOS #)
FILE11
FILE13
FILE15
FILE30
Temporary files used: none
Files generated: CHECK
FILE6
FILE12
D. Input format:
# INTCOS #
1. FORMAT(10I5)
(1) NST number of bond lengths
(2) NBND number of bond angles
(3) NLIB number of linear bond angles
(each 180 degree angle should be counted twice)
(4) NDEF number of out-of-plane angles
(5) NTORS number of torsional angles
(6) IFORCE = 0 ... no force constant matrix
= 1 ... calculate force constant matrix with respect
to internal coordinates (read in Cartesian
second derivatives from FILE15)
(7) IGRAD = 0 ... read geometry and gradient from FILE30
= n ... flag to allow different sets of gradients
and geometries to be read in
E.g., 1103 means 3rd set from FILE11
1304 means 4th set from FILE13
(8) IATOM number of atoms - required if IGRAD .NE. 0
(9) ICOOD number of internal coordinates - required if
IGRAD .NE. 0
(10) IPRNT = 0 ... minimum printing
= 1-6 .. more output
.....for more information about lines 2-6, see input description
for the program NORMCO.
2. FORMAT(2I5) input data for bond lengths
(KR(I),LR(I),I=1,NST)
3. FORMAT(3I5) input data for bond angles
(KA(I),LA(I),MA(I),I=1,NBND)
4. FORMAT(3I5) input data for linear bond angles
(KB(I),LB(I),MB(I),I=1,NLIB/2)
5. FORMAT(4I5) input data for out-of-plane angles
(KD(I),LD(I),MD(I),ND(I),I=1,NDEF)
6. FORMAT(4I5) input data for torsional angles
(KT(I),LT(I),MT(I),NT(I),I=1,NTORS)
________________________________________________________________________
1 45. GNEXTS
------
A. GNEXTS is a geometry optimization program based on the steepest
descent method.
This program is able to treat up to 50 atoms and 150 internal
coordinates.
B. References: none
C. Files required: INPUT (# GNEXTS #)
FILE12
FILE15 (if IHESS = 1)
FILE30
Temporary files used: none
Files updated: FILE13
FILE30
Files generated: CHECK
FILE6
D. Input format:
# GNEXTS #
1. FORMAT(8I5)
NCOOD number of symmetrically distinct internal
coordinates to be optimized
NSORT = 0 ... no sort - use last NCOOD+1 values in FILE13
= 1 ... total (global) gradient sort
= 2 ... energy sort
= 3 ... selected gradient sort
IHESS (parameter for Hessian matrix)
= 0 ... no Hessian matrix
= 1 ... use Hessian matrix for geometry optimization
IMETR (parameter for the variable metric method)
= 0 ... skip this method
= 1 ... Murtagh-Sargent method
= 2 ... Fletcher method
= 3 ... Davidon-Fletcher-Powell method
IAGHES (parameter for the augmented Hessian matrix
= 0 ... skip this method
= 1 ... use this method
NVAR number of Cartesian coordinates to be optimized
NROOT number of the root to be pursued (default = 1)
IPRNT = 0 ... minimum printing
= 1-6 .. more output
2. FORMAT(14I5)
(NSET(I),I=1,NCOOD)
the numbers of the symmetrically distinct internal
coordinates to be optimized as defined in INTCOS
For example, just list 1 and 3 if you defined three
coordinates in INTCOS but you only wish to optimize
coordinate number 1 and coordinate number 3.
3. If NVAR is not equal to 0:
FORMAT(14I5)
(NXVAR(I),I=1,NVAR)
NXVAR(I) is the number of the Cartesian coordinate
to be optimized. The coordinates are numbered as
1 = x1, 2 = y1, 3 = z1, 4 = x2, 5 = y2, etc.
4. FORMAT(A5,3I5)
UPDATE = UP update geometry in FILE30
= blank no update for FILE30
NCHNG 10 ** (-NCHNG) is the tolerance for geometry change
in each iteration (default --- 2)
NCONV 10 ** (-NCONV) is the convergence criterion for
geometry optimization (default --- 7)
NUNIQ number of unique atom sets (default --- NATOM)
5. If NUNIQ is not equal to NATOM:
FORMAT(14I5)
(NUNQ(I),(NAT(I,J),J=1,NUNQ(I)),I=1,NUNIQ)
where NUNQ(I) is the number of atoms in Ith set
NAT(I,J) is a list of atom numbers in Ith set
________________________________________________________________________
1 46. NEWTON
------
A. NEWTON is a geometry optimization program based on the
Newton-Raphson method.
This program is able to treat up to 50 atoms and 150 internal
coordinates
B. References: none
C. Files required: INPUT (# NEWTON #)
FILE12
FILE15 (if IHESS = 1)
FILE30
Temporary files used: none
Files updated: FILE13
FILE30
Files generated: CHECK
FILE6
D. Input format:
# NEWTON #
1. FORMAT(5I5)
NCOOD number of symmetrically distinct internal
coordinates to be optimized
NSORT = 0 ... no sort - use last NCOOD+1 values in FILE13
= 1 ... total (global) gradient sort
= 2 ... energy sort
= 3 ... selected gradient sort
IMETR (parameter for the variable metric method)
= 0 ... skip this method
= 1 ... Murtagh-Sargent method
= 2 ... Fletcher method
= 3 ... Davidon-Fletcher-Powell method
NVAR number of Cartesian coordinates to be optimized
IPRNT = 0 ... minimum printing
= 1-6 .. more output
2. FORMAT(14I5)
(NSET(I),I=1,NCOOD)
the numbers of the symmetrically distinct internal
coordinates to be optimized as defined in INTCOS
For example, just list 1 and 3 if you defined three
coordinates in INTCOS but you only wish to optimize
coordinate number 1 and coordinate number 3.
3. If NVAR is not equal to 0:
FORMAT(14I5)
(NXVAR(I),I=1,NVAR)
NXVAR(I) is the number of the Cartesian coordinate
to be optimized. The coordinates are numbered as
1 = x1, 2 = y1, 3 = z1, 4 = x2, 5 = y2, etc.
4. FORMAT(A5,3I5)
UPD = UP update geometry in FILE30
= blank no update for FILE30
NCHNG 10 ** (-NCHNG) is the tolerance for geometry change
in each iteration (default --- 2)
NCONV 10 ** (-NCONV) is the convergence criterion for
geometry optimization (default --- 7)
NUNIQ number of unique atom sets (default --- NATOM)
5. If NUNIQ is not equal to NATOM:
FORMAT(14I5)
(NUNQ(I),(NAT(I,J),J=1,NUNQ(I)),I=1,NUNIQ)
where NUNQ(I) is the number of atoms in Ith set
NAT(I,J) is a list of atom numbers in Ith set
________________________________________________________________________
1 47. BMWRTA
------
A. BMWRTA takes the geometry and gradients from the bottom of FILE11
and writes them into the BMAT file in the appropriate place
(i.e. after the option lines). The old Cartesian coordinates and
forces in the BMAT file are overwritten. Thus, the easiest way to
set up a virginal BMAT file is to leave 2N blank lines after the
options, where N is the number of atoms.
B. References: none
C. Files required: BMAT
FILE11
Temporary files used: none
Files updated: BMAT
Files generated: none
D. Input required: none
________________________________________________________________________
1 48. BMATIN6
-------
A. BMATIN6 is an extensively modified version of Prof. Peter Pulay's
"BMAT" program. BMATIN6 also incorporates the eigenvector following
routine OPTEFC of Dr. Jon Baker.
Notes on the B-matrix program:
The main tasks performed by BMATIN6 are:
1. Geometry optimizations can be carried out completely in
internal coordinates. BMATIN6 takes Cartesian geometries and
energy gradients and transforms these to internal coordinates and
internal forces. The user also supplies an initial-guess force
constant matrix in internal coordinates. This can be built up
from values found in the literature for previous calculations or
for experiments. Alternatively, a set of force constants from a
small basis set calculation can be used. The force constants are
updated using various Hessian update methods:
a) Powell (symmetric Broyden), which does not enforce positive
definiteness.
This is the default for transition structures.
b) Davidon-Fletcher-Powell
c) The variable metric method of Murtaugh and Sargent.
d) Broyden-Fletcher-Goldfarb-Shanno (BFGS), which is usually
the best for equilibrium geometries.
This is the default for minima.
New internal coordinates are obtained according to the Newton-
Raphson scheme and transformed back into Cartesian
coordinates. These are then appended to the bottom of the file
called INPUT in the format appropriate for the GEOMIU program.
2. Cartesian coordinates can be generated for displacements along
internal coordinates. BMATIN6 takes as input the reference
geometry in Cartesian coordinates along with the specification
of the internal coordinate distortion(s) to be made. The
Cartesian coordinates corresponding to the internal coordinate
distortion(s) are appended to the bottom of the file called INPUT.
If atomic masses are given, the transformations between Cartesian
and internal coordinates will be in accordance with the Eckart
conditions (see Wilson, Decius & Cross "Molecular Vibrations"
(1955), Section 11-1). (This is important if displacements of
dipole moments are desired.)
3. BMATIN6 can be used to simply compute and print the B-matrix
which transforms Cartesian coordinates to internal coordinates.
Notes:
a. Geometry optimizations in internal coordinates are always
carried out in the totally symmetric irreducible
representation of the point group of the system in question
(e.g. A' for Cs, A1 for C2v, Ag for D2h, etc.). Some of
the most frequent causes of error in the use of BMATIN6 are
the attempt to use fewer coordinates than there are degrees
of freedom, and the use of redundant sets of coordinates.
For H2O, it is obvious that there are two A1 coordinates that
one would use to optimize the molecule. For very large
systems, however, it is sometimes tedious to work out the
number of coordinates of a particular symmetry and some useful
time-saving formulas are given in H. H. Jaffe and M. Orchin
"Symmetry in Chemistry" (Wiley-Interscience, New York, 1965),
Appendix 2.
If it is desired to obtain asymmetric displacements then all
3N-6 internal coordinates need to be specified.
b. If an optimization gets bollixed up in any way, it may be that
the file called RESUL2 is causing problems. If the old
geometry in RESUL2 and the geometry in BMAT are the same, the
Hessian updates will behave badly. It may be necessary to
erase the RESUL2 file and start the update over.
c. Occasionally problems may arise due to a discontinuity in the
definition of an internal coordinate (e.g. a nearly linear
angle or out-of-plane angle near 90 degrees). Solution: try
different angle definitions.
Notes on the eigenvector following routine (invoked with the EIGF
option):
The eigenvector following routine (OPTEFC) is an efficient
quasi-Newton algorithm for locating transition structures. It
was written by Jon Baker as a local addition to the GAUSSIAN 82
package in Leo Radom's group.
The method is based on a modification to the Newton-Raphson
step first proposed by Cerjan and Miller <1>, although the major
part of the algorithm is based on the later developments of Simons
and coworkers <2,3>. It is capable of locating transition
structures even if started in the wrong region of the energy
surface, and, by invoking Hessian mode following, can locate
several different transition structures from the same initial
starting point.
It can also be used to locate minima.
A discussion of the formalism and the ideas behind it,
together with a description of the algorithm and some practical
examples are given in ref <4>.
Mode following:
For a transition structure search, maximization normally
takes place along the lowest mode and minimization along all other
modes. However, as first pointed out by Cerjan and Miller <1>,
it is possible to maximize along modes other than the lowest and
in this manner obtain transition structures for alternative
rearrangements and/or dissociations from the same initial
starting point.
Mode following is switched on for OPTEFC by means of the MODE
option in BMAT. "MODE 1 n" for a particular variable "n"
will cause a transition structure search to follow the Hessian
mode with the largest magnitude component for that variable.
The idea behind this is that, in many cases, the various Hessian
modes are dominated by a single variable corresponding to a large
change in a particular bond length or bond angle say, and it is
this particular parameter that is required to change the most
during a transition structure search. For example, looking for a
dissociation transition structure should involve change in
essentially just one parameter - the bond length between the two
dissociating moieties - and following the mode with the largest
magnitude component for this bond length should have the best
chance of leading to the desired transition structure.
This is not always the case however, and specific Hessian
modes can be followed without any reference to particular
internal variables by using "MODE 0 n", causing the nth
mode to be followed.
Note that only one mode can be followed at a time.
Error messages and program limitations in OPTEFC:
A maximum of 50 variables can be specified.
A certain amount of input checking is done in the subroutine
INITEF and most of the error message printout occurs here.
Nothing else should go wrong, although it is theoretically
possible for the iterative procedure which calculates the
eigenvalue shift parameter lambda in subroutine FORMD to fail.
Either the procedure will not converge, in which case the message
****************************************
** UNABLE to determine lamda in FORMD **
****************************************
will be printed out, or convergence will be attained, but to an
unacceptable value, giving the message
*****************************************
** ERROR in determining lamda in FORMD **
*****************************************
It is EXTREMELY UNLIKELY for either of these events to occur.
If they do, the internal coordinates should be checked carefully;
specifying more variables than are allowed for by symmetry may be
what is causing the problem. Also double check that the geometry
and forces in the BMAT file have been updated using BMWRTA and
that they are different to the "old" ones in RESUL2.
Note that switching on the Newton-Raphson step (setting IOP19=1)
obviates the need to calculate a shift parameter, although this
can only be used in the right region of the energy surface.
B. References:
BMAT:
P. Pulay in "Applications of Electronic Structure Theory", ed. H.
F. Schaefer (Plenum, New York, 1977) p. 165.
OPTEFC
<1> C. J. Cerjan and W. H. Miller, J. Chem. Phys. 75 (1981) 2800.
<2> J. Simons, P. Jorgensen, H. Taylor and J. Ozment, J. Phys.
Chem. 87 (1983) 2745.
<3> A. Banerjee, N. Adams, J. Simons and R. Shepard, J. Phys. Chem.
89 (1985) 52.
<4> J. Baker, J. Comput. Chem. 7 (1986) 385.
C. Files required: BMAT
RESUL2 Note: on the first optimization
cycle, RESUL2 should contain
just one blank line, 132
characters long.
Temporary files used: none
Files generated: RESUL1 (main output)
RESUL3 (internal forces)
MAKEFT (cummulative RESUL3)
Files updated: RESUL2
INPUT
D. Input format:
The input for this program is read from a file called BMAT.
The format style is different from the other programs in the PSI
package.
1. The first line should contain the following (beginning in the
first column):
BMAT
The rest of this first line would normally contain information
meaningful to the user.
2. FORMAT(A4,4X,I2)
The second line should contain
CARD n
where n is the number of atoms.
Lines 3 to 23 are optional and may be in any order.
FORMAT(A4,1X,I5) unless otherwise specified.
3. ANGS
This line indicates that the nuclear coordinates are to be
read in Angstrom units as opposed to atomic units (default).
4. FIXC n
If this option is specified, internal coordinate "n" is fixed
in the geometry optimization. To fix more than one coordinate
repeat this line as many times as necessary.
5. PUNC
This causes the program to print the B-matrix to the file
called RESUL3.
6. PRIN
If this option is present, the input for the specification
of the internal coordinates is printed to the main output file
(called RESUL1). The values of the simple internal
coordinates are also printed.
7. GDYN
With this option, the Cartesian forces are read in with units
of mdyne. Keep in mind that the forces in FILE11 are in
atomic units, so this option should only be used if the forces
are to be typed in by hand.
8. FINT
If this option is specified, the program reads in internal
forces instead of Cartesian forces. Note the different place
where the internal forces are read in: after the specification
of the internal coordinates. The Cartesian forces are read in
before the internal coordinates. The dimensions of the
internal forces should be compatible with the energy measured
in AJ (=mdyne*A) and with the stretching coordinates measured
in Angstrom, bending ones in radian.
9. FMAT
This line indicates that a force constant matrix is to be read
in. The units of the force constants are mdyne/A, mdyne or
mdyne*A depending on the type of internal coordinate.
If this is the first cycle, the force constant matrix is read
in from the BMAT file (see below).
On subsequent cycles, the force constant matrix and old data
(internal coordinates, forces, and displacements) are read in
from the file called RESUL2.
The program keeps track of the number of optimization cycles
through the information on the file called RESUL2. On the
first cycle, RESUL2 should contain just one blank line, 132
characters long.
10. NOEX
This option supresses the reading in of any information from
the file called RESUL2, and the data is read in from the
BMAT file instead.
11. NOUP
If this option is specified, Hessian updating is not
performed. Otherwise, the internal coordinates and forces
in the previous step of the geometry optimization are used
to improve the force constant matrix using one of the
following methods:
12. MURT
Murtagh-Sargent update.
13. POWL
Powell update. (Default for transition structures)
14. DFLP
Davidon-Fletcher-Powell update.
15. BFGS
Broyden-Fletcher-Goldfarb-Shanno update. (Default for minimum)
16. FORMAT(A4,6X,4F6.0)
FLT1, ETA1, ETA2, ETA3, ETA4
FLT1 specifies that the Fletcher-Powell method of optimization
is to be used rather than the variable metric method of
Murtagh-Sargent. With FLT1, the first part of the algorithm
is implemented. The FMAT option must also be specified if the
FLT1 option is present. The ETAn values are steps along the
Fletcher-Powell direction vector for which Cartesian
coordinates are desired. If, for example, only two steps are
desired, leave ETA3 and ETA4 blank. The new sets of Cartesian
coordinates are appended to the bottom of the INPUT file in
the format appropriate for the GEOMIU program.
17. FORMAT(A4,8X,3(F6.0,F16.10))
FLT2, ETA1, ENERGY1, ETA2, ENERGY2, ETA3, ENERGY3
With FLT2 the second part of the Fletcher-Powell algorithm is
implemented. The three (ETA,ENERGY) pairs allow computation
of the ETA value which minimizes the energy along the
Fletcher-Powell vector, whence Cartesian coordinates for a
new gradient calculation are computed.
18. FORMAT(A4,1X,9I5)
EIGF, IOP5, IOP7, IOP8, IOP13, IOP16, IOP17, IOP19, IOP33, IOP34
The EIGF option specifies that the eigenvector following
routine OPTEFC is to be used to perform the geometry
optimization.
IOP5 Nature of required stationary point
= 0 ... Find a TS (DEFAULT)
= 1 ... Find a minimum
IOP7 not used at present
IOP8 Maximum stepsize allowed during optimization
= 0 ... DMAX = 0.3 (DEFAULT)
= n ... DMAX = 0.01*n
IOP13 Type of Hessian update
= 0 ... Powell update (DEFAULT)
= 1 ... BFGS update (used for minima)
= 2 ... BFGS update with safeguards to ensure retention
of positive definiteness
IOP16 Maximum allowable magnitude of Hessian eignvalues
If this magnitude is exceeded, the eigenvalue is
replaced.
= 0 ... EIGMAX = 25.0 (DEFAULT)
= n ... EIGMAX = 0.1*n
IOP17 Minimum allowable magnitude of Hessian eigenvalues
Similar to IOP16.
= 0 ... EIGMIN = 0.0001 (DEFAULT)
= n ... EIGMIN = 1.0/n
IOP19 Search selection
= 0 ... P-RFO or RFO step only (DEFAULT)
= 1 ... P-RFO or RFO step for "wrong" Hessian otherwise
Newton-Raphson
IOP33 Print option
= 0 ... ON (DEFAULT)
= 1 ... OFF turns off extra printing
(Default of "ON" by popular request)
IOP34 Dump option
= 0 ... OFF (DEFAULT)
= 1 ... ON turns on debug printing
Note : Setting IOP13 = 2 ensures that the BFGS update
(the default update for a minimum search) retains positive
definiteness; i.e. if the Hessian before the update has all
positive eigenvalues, then so will the updated Hessian.
In most cases the BFGS update retains positive definiteness
anyway, but this is not guaranteed. Use of this option will
cause the update to be skipped if positive definiteness is
endangered. Thus, once the Hessian becomes positive definite,
it will remain so within the limits of numerical rounding
error. Such a feature is, of course, not desirable for a
transition structure search, and use of the BFGS update is
consequently not recommended when searching for a transition
structure.
If there is a conflict in the updating methods specified
by IOP13 and EXPLICITLY by BMAT, the BMAT option will take
precedence.
19. FORMAT(A4,1X,2I5)
MODE n m
This option turns on mode following in OPTEFC.
If the first number is 0 (default), the second is the
number of the Hessian mode (as ordered by eigenvalue) to be
followed. In transition structure searches, the second number
is 1 by default (i.e. the lowest mode).
If the first number is 1, the second is the number of the
internal coordinate that determines which mode is followed
(the mode with the highest component for that internal
coordinate).
20. DISP n
This line indicates that the aim of the calculation is not to
transform forces but to obtain molecular geometries which are
distorted in a prescribed way from the reference geometry.
"n" is the number of displacements. Note that the distortions
are exact curvilinear distortions.
Default units for displacements are Angstrom and radians.
For DISP to work correctly, a dummy RESUL2 file is required
containing just one blank line, 132 chacters long.
21. BOHR
This option is for use with the DISP option. It specifies
that displacements for bond stretching coordinates are given
in bohr rather than Angstrom. (Units for angular coordinates
are still in radians.)
22. DEGR
This option is for use with the DISP option. It specifies
that displacements for angular coordinates are given in
degrees rather than radians. (Units for stretching
coordinates are still in Angstrom.)
23. FORMAT(A4,12X,3F16.10)
DUMB, X, Y, Z
This option is for use with the DISP option. It specifies
the Cartesian coordinates of a dummy atom for linear bends.
The program zooms the dummy atom out to a distance of 1
billion angstrom so that displacements using LIN1 and LIN2
are degenerate. The Cartesian coordinates generated do not
include the dummy atom.
24. FORMAT(1X,A5,2X,I2,6X,3F16.12,F10.5)
SYMB(I) atomic symbol (used for print out only)
IA(I) atomic number (used for print out only)
X(I) x coordinate
Y(I) y coordinate
Z(I) z coordinate
M(I) atomic mass in atomic mass units (optional)
.....repeat this line for each atom
Don't worry about this section too much. If you are doing
an optimization, just leave N blank lines (where N is the
number of atoms) and let BMWRTA take care of the format.
If the masses are specified, the calculation of the new
Cartesian coordinates is performed such that the Eckart
conditions are obeyed. It is important to do this if you
wish to calculate displacements of dipole moments.
25. If the DISP option is not specified and neither is the FINT
option, then the Cartesian gradients must be given here
in hartree/bohr:
FORMAT(3F16.12)
FX(I) x coordinate gradient
FY(I) y coordinate gradient
FZ(I) z coordinate gradient
.....repeat this line for each atom
If you are performing an optimization, just leave N blank
lines (where N is the number of atoms) and let BMWRTA take
care of the format.
26. If the DISP option is on then the definition of the distortions
from the reference geometry must be given here in internal
coordinates:
FORMAT(4(I2,2X,F12.8,2X))
I1, D1, I2, D2, I3, D3, I4, D4
.....repeat this line for each displacement
Coordinate I1 is distorted by D1, coordinate I2 by D2, and so
on. Some of the Is may of course be zero (leave them blank).
The (default) units of displacement are Angstroms and radians.
Note that a simultaneous displacement takes place for all four
coordinates. It is not possible to specify a displacement
which affects more than four coordinates simultaneously.
However, one can change four coordinates, then take the result
as a new reference geometry and change four others and so on.
27. Definition of internal coordinates:
FORMAT(A4,6X,F16.12,A4,6X,5I3)
KW, COEFF, TYPE, A, B, C, D
.....repeat this line until all coordinates are defined
KW may be either the character "K" (in column 1) or
blank.
If it is "K", this shows that the present coordinate
begins a new, independent internal coordinate.
If it is blank, the coordinate is interpreted as the
continuation of a composite coordinate begun earlier.
Any other character in columns 1-4 terminates the
input of the internal coordinates.
COEFF is the coefficient of the simple internal coordinate
in the linear combination for composite internal
coordinates. Zero or blank is interpreted as 1.0.
The coefficients are normalized by the program.
TYPE = STRE for bond stretching coordinates
= INVR for inverse bond length coordinates
= BEND for bond angle coordinates
= OUT for out-of-plane coordinates
= TORS for torsion coordinates
= LIN1 for the deformation of a linear chain of atoms in
the plane of a fourth atom
= LIN2 is like LIN1, but the deformation is perpendicular
to the plane of the four atoms
A-D are the numbers of the nuclei that take part in the
coordinate.
The internal coordinates are defined as follows:
For STRE, the coordinate is the A-B bond distance, and the order
of A and B does not matter.
For INVR, the coordinate is the A-B bond inverse, and the order
of A and B does not matter.
For BEND, it is the A-B-C bond angle. A and C can be exchanged
but the central atom must be B.
For OUT, the coordinate is the angle between the AB vector and
the plane containing the angle C-B-D. The coordinate is
positive if A is on the same side of the plane as the
vector BC X BD. Note that the central atom comes second
here and that C and D can be exchanged but that this
changes the sign of the coordinate.
For TORS, the coordinate is defined as the angle between the
planes ABC and BCD. Note that ABCD and DCBA are
equivalent.
For LIN1, the coordinate is the collinear bending of the linear
chain of atoms ABC in the the plane which contains D. The
sign is positive if A and C move towards D.
For LIN2, the coordinate is the bending of ABC perpendicular to
the plane which contains D. The sign is positive if A and
C move in the direction of the vector product BD X BA.
28. If the FINT option is specified:
FORMAT(F16.12)
F(I) internal force
.....repeat this line for each internal coordinate
29. If the FMAT option is present and NOEX is not specified (or it
is the first optimization cycle), then an approximation to
the internal coordinate force constant matrix is read in:
FORMAT(8F10.7)
((FC(I,J), J=1,I), I=1,NQ) where NQ is the number of
internal coordinates
Each row of the force constant matrix is read up to and
including the diagonal element. Each row begins on a new
line.
Don't be put off by the requirement of a force constant
matrix. If you know nothing about the system being studied,
just use values of 3.0 - 8.0 (for stretching coordinates) and
1.0 (for bending coordinates) for the diagonal force
constants, and leave the rest zero. If experimental force
constants are used, they should be scaled by 1.1 .
For transition structure optimizations, it is important to
start with an analytical Hessian (calculated, perhaps, at a
very low level of theory).
30. If the NOEX option is specified without NOUP, the information
for the update must be read in here (unless this is the first
cycle, which needs no data for the update):
FORMAT(3F16.12)
The information required is the old internal coordinates, the
old forces, and the displacements to make the present internal
coordinates.
31. FORMAT(A4)
STOP
This line terminates the input to BMATIN6.
________________________________________________________________________
1 49. DIPDER
------
A. DIPDER calculates derivatives of dipole moments.
B. References:
Y. Yamaguchi, M. Frisch, J. Gaw, H. F. Schaefer and J. S.
Binkley, J. Chem. Phys. 84 (1986) 2262.
C. Files required: INPUT (# DIPDER #)
FILE30
Temporary files used: none
Files generated: CHECK
FILE6
FILE43
D. Input format:
# DIPDER #
1. FORMAT(I5)
IPRNT = 0 ... minimum printing
= 1-6 .. more output
________________________________________________________________________
1 50. RAMANC
------
A. RAMANC calculates the MO contributions to the electric
polarizability derivatives for closed shell SCF wavefunctions
and writes them to FILE18.
RAMANC should only be used for non-degenerate closed shell systems.
B. References:
C. Files required: INPUT (# RAMAN ##)
FILE40
FILE43
FILE44
Temporary files used: none
Files generated: CHECK
FILE6
FILE18
D. Input format:
# RAMAN ##
1. FORMAT(I5)
IPRNT = 0 ... normal printing
= 1-4 .. more output
________________________________________________________________________
1 51. RAMINT
------
A. RAMINT calculates the AO contributions to the electric
polarizability derivatives for SCF wavefunctions. The total
polarizability derivatives are then written to FILE18.
B. References:
C. Files required: INPUT (# RAMINT #)
FILE18
FILE30
FILE40
FILE44
Temporary files used: none
Files updated: FILE18
Files generated: CHECK
FILE6
D. Input format:
# RAMINT #
1. FORMAT(2(A8,2X))
SCFTYP = CLSCF
= GRSCF
= TCSCF * not available at present
= MCSCF * not available at present
CITYP = SCF
2. FORMAT(I5)
IPRNT = 0 ... normal printing
= 1-4 .. more output
3. FORMAT(A8)
NOSYM = NOSYM symmetry turned off
This MUST be used.
________________________________________________________________________
1 52. PROPER
------
A. PROPER performs a Mulliken population analysis and calculates the
dipole moments of the SCF wavefunction.
B. References:
R. S. Mulliken, J. Chim. Phys. 46 (1949) 497, 675.
R. S. Mulliken, J. Chem. Phys. 23 (1955) 1833, 1841.
R. S. Mulliken, J. Chem. Phys. 36 (1962) 3428.
C. Files required: INPUT (# PROPER #)
FILE30
Temporary files used: none
Files generated: CHECK
FILE6
D. Input format:
# PROPER # (default values used if # PROPER # is not found)
1. FORMAT(3I5)
ISCFCI not used in this version
ICENT = 0 or 1 (default = 1) for center of mass as the
reference coordinate of dipole moment
= 2 ... for origin of space fixed coordinate
= 3 ... for center of charge based on Mulliken
population
= 4 ... for center of nuclear charge
= 5 ... for center of net charge
Values 2-5 may be used for charged systems
(for which cases the dipole moment
definition is ambiguous).
IPRINT = 0 ... minimum printing
= 1-6 .. more output
________________________________________________________________________
1 53. CIPROP
------
A. CIPROP performs a Mulliken population analysis of the
CI or CC wavefunction and calculates the CI or CC dipole moment
(without the CPHF correction).
B. References: see PROPER
C. Files required: INPUT (# PROPER #)
FILE30
FILE40
Temporary files used: none
Files generated: CHECK
FILE6
FILE59
D. Input format:
# PROPER # (default values used if # PROPER # is not found)
1. FORMAT(3I5)
ISCFCI should be set equal to 0 or 2
ICENT = 0 or 1 (default = 1) for center of mass as the
reference coordinate of dipole moment
= 2 ... for origin of space fixed coordinate
= 3 ... for center of charge based on Mulliken
population
= 4 ... for center of nuclear charge
= 5 ... for center of net charge
Values 2-5 may be used for charged systems
(for which cases the dipole moment
definition is ambiguous).
IPRINT = 0 ... minimum printing
= 1-6 .. more output
________________________________________________________________________
1 54. BONDEX
------
A. BONDEX calculates bond orders and valencies (both Mulliken and
Lowdin). The program is limited to 200 basis functions and 112
atoms.
B. References:
See PROPER, and:
P. O. Lowdin, J. Chem. Phys. 18 (1950) 365.
C. Files required: INPUT (# BONDEX #)
FILE30
FILE40 (if ISCFCI = 2)
Temporary files used: none
Files generated: CHECK
FILE6
D. Input format:
# BONDEX #
1. FORMAT(3I5)
ISCFCI = 0 or 1 SCF wavefunction
= 2 ... CI or CC wavefunction
IPRNT = 0 or 1 minimum printing
= 2 ... more output, including density matrix
ITYFC = 0 or 3 s p d functions only (default)
= 5 ... f and g function are present
If using ISCFCI = 2, BONDEX would normally be run after the
program LAGTR. Alternatively, it is possible to calculate
just the CI energy and then run MASTER, ONEPDM (with the
options PRINT = 1 and PRPFLG >= 1) and BONDEX.
________________________________________________________________________
1 55. NORMCO
------
A. NORMCO transforms the Cartesian second derivatives into normal
coordinates and performs a vibrational frequency analysis.
B. References:
C. Files required: INPUT (# NORMCO #)
FILE30
FILE15
FILE17 (if IDIPOL = 1)
FILE18 (if IPOLAR = 1)
Temporary files used: none
Files generated: CHECK
FILE6
D. Input format:
# NORMCO #
1. FORMAT(12I5)
IFXGF = 0 or 1 FX matrix method
= 2 ... GF matrix method
ISOTOP = 0 or 1 use regular atomic masses
= 2 or more number of isotopomers
NVIB vibrational degrees of freedom |
IFORCE = 1 ... read in gradient sets from FILE30 and FILE15 |
(for finite difference force constants) |
= 2 ... read in force constants from INPUT file |
= 3 ... read in a Hessian from FILE15 |
(analytical force constants in terms of |
Cartesian coordinates) |
IDIPOL = 0 ... no dipole derivatives
= 1 ... read in dipole derivatives (FILE17)
IPOLAR = 0 ... no polarizability derivatives
= 1 ... read in polarizability derivatives (FILE18)
IGEOMT = 0 ... read in geometry from FILE30
= 1 ... read in geometry from INPUT file
ITHERM = 0 or 1 symmetry number = 1
= n ... symmetry number = n
this is the sigma used for the rotational partition
function
IQELEC = 0 or 1 degeneracy of ground electronic state = 1
= n ... degeneracy of ground electronic state = n
IZVLIM threshold to include frequencies in ZPVE calculation
= 0 ... threshold = 20 cm-1
= n ... threshold = n cm-1
IPLOT not used
IPRNT = 0 ... normal printing
= 1-6 .. more output
2. If IGEOMT = 1:
FORMAT(4F20.10)
COORD(1,I) x coordinate
COORD(2,I) y coordinate
COORD(3,I) z coordinate
W(I) atomic mass
.....repeat this line for each atom
2A. If IFORCE = 1: |
FORMAT(F10.7) |
DELX perturbation of Cartesian coordinate for |
finite difference method |
|
2B. If IFORCE = 2: |
FORMAT(3F20.10) |
(FX(I,J),J=1,N3N),I=1,N3N) |
force constants (N3N = 3 times the number |
of atoms) |
|
3. If IFXGF = 2:
FORMAT(6I5) parameters for GF matrix method
NST number of stretching coordinates
NBND number of bending coordinates
NLIB number of linear bending coordinates
(each 180 degree angle should be counted twice)
NDEF number of out-of-plane coordinates
NTORS number of torsional coordinates
ISYM = 0 ... no symmetry internal coordinates
= 1 ... use symmetry internal coordinates
.....lines 6-10 are parameters for simple internal coordinates
4. If IFXGF = 2:
FORMAT(2I5)
KR(I),LR(I) numbers of atoms for each stretch
.....repeat this line NST times (i.e. I=1,NST)
5. If IFXGF = 2:
FORMAT(3I5)
KA(I),LA(I),MA(I) numbers of atoms for each bend for the
bond angle KA-LA-MA
.....repeat this line NBND times (i.e. I=1,NBND)
6. If IFXGF = 2:
FORMAT(3I5)
KB(I),LB(I),MB(I) numbers of atoms for each linear bend
for the 180 degree bond angle KB-LB-MB
.....repeat this line NLIB/2 times (i.e. I=1,NLIB/2)
7. If IFXGF = 2:
FORMAT(4I5)
KD(I),LD(I),MD(I),ND(I) numbers of atoms for each out-of-plane
angle (ND is the apical atom)
If PI = KD out of the LD-ND-MD plane and
THETA = the LD-ND-MD angle, then the
out-of-plane angle = PI * sin(THETA).
(Note: this is different to BMATIN6 and
INTDER, which define the out-of-plane angle
to be PI.)
.....repeat this line NDEF times (i.e. I=1,NDEF)
8. If IFXGF = 2:
FORMAT 4I5
KT(I),LT(I),MT(I),NT(I) numbers of atoms for each torsion for
the dihedral angle between planes
KT-LT-MT and LT-MT-NT
.....repeat this line NTORS times (i.e. I=1,NTORS)
9. If IFXGF = 2 and ISYM = 1:
FORMAT(4I5)
II,JJ,KK,LL transformation parameters for symmetry
coordinates
II: series number of a symmetry coordinate
JJ: one of the component simple coordinates
for II
KK: numerator of the coeficient for JJ in II
LL: square of denominator of the coeffecient
for JJ in II
.....There may be several lines for the same II if it is made up of
more than one component simple coordinate.
.....Repeat line 11 as many times as you need, and then put four
zeros (4I5) to terminate the input of the symmetry
coordinates.
10. If ISOTOP is greater than 1:
FORMAT(F20.10)
WISO(I) isotopic atomic mass
.....repeat this line for each atom
.....section 12. should occur ISOTOP-1 times. The first normal
coordinate analysis is always performed with regular atomic
masses (unless IGEOMT = 1, in which case it uses the masses
from section 2. above).
________________________________________________________________________
1 56. INTDER or NINTDER
------ -------
A. INTDER performs general curvilinear transformations among higher
order derivatives (Cartesian <---> internal) and may be used to
calculate vibrational frequency analyses in either internal or
Cartesian coordinates. The internal coordinates used may be either
simple coordinates or symmetrized combinations.
Notes:
It is not necessary to use all 3N-6 internal coordinates when
doing the transformations, but it is wise to transform
complete symmetry blocks (i.e. work out how many normal modes
there are of a particular symmetry and then you need to have
the same number of non-redundant symmetry internal
coordinates).
B. References: none
C. Files required: INTDER1 (# INTDER #)
FILE11 (if NGEOM = 0)
FILE15 (if NINV = 0 and NDER = 2)
FILE17 (if NINV = 0 and NVEC = 1 or
if NFREQ= 1 and IRINT= 1)
FILE20 (if NINV = 0 and NDER = 3)
FILE24 (if NINV = 0 and NDER = 4)
FILE12 (if NINV = 1 and NDER = 1)
FILE16 (if NINV = 1 and NDER = 2)
FILE18 (if NINV = 1 and NVEC = 1 or
if NFREQ= 1 and IRINT= 1)
FILE21 (if NINV = 1 and NDER = 3)
FILE25 (if NINV = 1 and NDER = 4)
11, 15, 20 and 24 are the 1st, 2nd, 3rd and 4th derivatives,
respectively, in Cartesian coordinates.
12, 16, 21 and 25 are the 1st, 2nd, 3rd and 4th derivatives,
respectively, in internal coordinates.
17 is the dipole moment derivatives in Cartesian coordinates.
18 is the dipole moment derivatives in internal coordinates.
Temporary files used: FILE91
FILE92
FILE93
FILE94
FILE95
FILE96
FILE97
Files generated: CHECK
INTDERO
FILE11 (if NINV = 1,2 and NDER = 1)
FILE15 (if NINV = 1,2 and NDER = 2)
FILE17 (if NINV = 1 and NVEC = 1)
FILE20 (if NINV = 1,2 and NDER = 3)
FILE24 (if NINV = 1,2 and NDER = 4)
FILE12 (if NINV = 0 and NDER = 1)
FILE16 (if NINV = 0 and NDER = 2)
FILE18 (if NINV = 0 and NVEC = 1)
FILE21 (if NINV = 0 and NDER = 3)
FILE25 (if NINV = 0 and NDER = 4)
D. Input format:
The input for this program is read from a file called INTDER1, which
has the same format as INPUT. The first section in this file should
be the # FILES ## input to tell the program what the temporary files
will be called. This is followed by:
# INTDER ##
1. FORMAT(16I5)
(1) NA number of atoms
(2) NS number of simple internal coordinates
(3) NSYM number of symmetry internal coordinates
(4) NDER highest order of derivative to be transformed
(1 to 4)
(5) NEQ = 0 ... no first derivatives transformed
(mainly used for stationary points)
= 1 ... first derivatives transformed
(6) NPRT a print option
(a four digit number which will explained below)
(7) NINV = 0 ... transform Cartesian derivatives to internal
coordinate derivatives
= 1 ... transform internal coordinate derivatives to
Cartesian derivatives
= 2 ... the same as = 1 except that the internal coordinate
derivatives are input from the INTDER input file
(see below)
(8) NDUM number of dummy atoms. Only used for the specification
of linear bending angles (LIN1). See (10) for reading
dummy atoms from either FILE11 or INTDER1.
(9) NTEST = 0 ... no test
= 1 ... numerically test and check the analytic
SR(I,J) and X(M,N) matrices
= -1 ... form the SR(I,J) and X(M,N) matrices
numerically and use these numerically
computed matrices in the transformation of
derivatives
= 2 ... numerically test and check the analytic
SR(I,J,K) and X(M,N,P) matrices
= -2 ... form the SR(I,J,K) and X(M,N,P) matrices
numerically and use these numerically computed
matrices in the transformation of derivatives
Numerical testing of derivatives of the internal coodinates
with respect to the Cartesian coordinates is useful for
debugging new types of coordinates added to the program
(10) NGEOM = 0 ... read Cartesian geometry from FILE11
(be sure the dummy atoms, if any, are
included in FILE11, but only in the geometry,
not in the first derivatives)
The geometry will be read from the bottom of
FILE11 unless the MULTI option is > 0.
= 1 ... read Cartesian geometry from the INTDER input file
(Be sure MULTI = 0,1 if run the program NINTDER.)
(11) NFREQ = 0 ... no frequency analysis performed
= 1 ... perform a frequency analysis in internal coordinates
= 2 ... perform a frequency analysis in Cartesian
coordinates
= 3 ... do both 1 and 2
= 4 ... the same as = 0 except that the force constants
are input from the INTDER input file (see below)
< 0 ... skip transformation of derivatives and just do
frequency analysis
(12) IRINT = 0 ... no IR intensities computed
= 1 ... IR intensities computed
Internal coordinate dipole moment derivatives
are read in from FILE18.
Cartesian coordinate dipole moment derivatives
are read in from FILE17.
= 2 ... same as 1, except that internal coordinate
dipole moment derivatives are read in from
the INTDER input file (see below)
set IRINT = 0 if NFREQ = 0
(13) NVEC = 0 ... no dipole moment derivatives transformed
= 1 ... dipole moment derivatives transformed
Masses are read in later so that the
transformation is performed according to the
Eckart conditions.
The dipole derivatives are read from FILE17
or FILE18 (see below for a description of
the input required).
(also set NDER = 1 and NINV = 0 or 1)
(NVEC = 1 assigns NEQ = 1)
It is not possible to transform dipole moment
derivatives at the same time as energy derivatives.
If NVEC = 1, NFREQ and IRINT should be equal to 0.
(14) NSTOP = 0 ... normal run
= 1 ... stop after forming the SR(I,J),X(M,N),SR(I,J,K),
and Y(M,N,P) matrices (as governed by NDER + NEQ).
(no auxiliary files are required (unless NGEOM = 0).
(15) MULTI = 0 ... normal run
= 1 ... normal run, except that the geometry (and
gradient) are read from the TOP of FILE11.
= n ... number of transformations to be done in one
run (e.g. set MULTI = n if there are "n" sets
of derivatives in FILE11 to be transformed).
If MULTI > 1, NGEOM must be 0 for the program
NINTDER.
2. FORMAT(A5,4I5,5X,A5)
TYPE(J), A, B, C, D, NUMST
TYPE(J) = STRE ... A-B bond length
= BEND ... A-B-C bond angle
= LIN1 ... A-B-C linear bond angle
Atom D is specified such that the vector
B-D is perpendicular to the bending plane.
Atom D is typically a dummy atom.
(Note: the definition of LIN1 given here
is equivalent to that of LIN2 given in the
BMATIN6 program.)
= OUT ... A out of the plane C-B-D
(i.e. the angle between the vector A-B and
the plane C-B-D)
The sign convention is the same as in BMATIN6.
(In the present version of this program,
OUT can be used only if NDER + NEQ < 3
(or NDER + NEQ < 4 if NTEST = -1).)
= TORS ... A-B-C-D torsion
(i.e. the angle between planes A-B-C and B-C-D)
= SPF ... Simons-Parr-Finlan coordinate for A-B
bond length i.e. (r-r0)/r
(If C=0, the bond length A-B is taken as
the reference.
If not, an additional card is read
immediately in F20.10 format defining the
reference bond length.)
A, B, C, D are integers defining the atoms involved
in the definition of internal coordinates
(if fewer than four integers required, set the
remaining to zero or leave them blank)
NUMTST = blank ... (default) If NTEST not = 0, all
coordinates are tested.
= ST ... use this to supress testing of individual
coordinates
.....repeat this line NS times (each line is for one internal
coordinate)
3. If NSYM > 0:
FORMAT(I5,4(I4,F14.10))
L, (IR(K), XR(K), K=1,4)
L symmetry coordinate number
IR simple internal coordinate number involved
XR coefficient of IR in L definition
All coefficients are automatically normalized in
the program.
When more than 4 simple internal coordinates are needed to
define one symmetry coordinate, use several lines with the
same L value.
A maximum of 5 lines can be used for one symmetry coordinate.
.....repeat line 3 until all symmetry coordinates are defined
Exit this section with L=0.
4. If NGEOM = 1:
FORMAT(3F20.10)
(XA(I,J),J=1,3) Cartesian geometry (x, y, z) in bohr
.....repeat this line NA+NDUM times (i.e. I = 1, NA+NDUM)
5. If NFREQ not = 0 or NVEC = 1:
FORMAT(F12.6)
XMASS(I) atomic masses in a.m.u.
.....repeat this line NA times (i.e. I = 1, NA)
6. If NINV = 2 read in the unique internal coordinate derivatives
which are non-zero. Use units consistent with the energy in
mdyne*Angstrom.
If NEQ not = 0:
FORMAT(I5,15X,F20.10)
M, F1(M)
End first derivatives with M=0.
If NDER >= 2:
FORMAT(2I5,10X,F20.10)
M, N, F2(M,N) M >= N is required.
End second derivatives with M=0.
If NDER >= 3:
FORMAT(3I5,5X,F20.10)
M, N, P, F2(M,N,P) M >= N >= P is required.
End third derivatives with M=0.
If NDER >= 4:
FORMAT(4I5,F20.10)
M, N, P, Q, F2(M,N,P,Q) M >= N >= P >= Q is required.
End fourth derivatives with M=0.
The format used here is consistent with that used in the file
called IDER which is produced by the program INTDIF.
Thus it is relatively straightforward to copy the derivatives
from IDER into the appropriate place in the INTDER1 file.
7. If NFREQ = +4 or -4:
FORMAT(7F10.6)
(F2(M,N), N=M,NSX) quadratic force constants in units
consistent with the energy in mdyne*A.
NSX = NSYM. If NSYM = 0 then NSX = NS.
.....repeat this line NSX times (i.e. M = 1, NSX)
8. If IRINT = 2:
FORMAT(3F20.10)
(U(I,J), J=1,3) internal (symmetry) coordinate dipole
moment derivatives (x, y, z) in units of
Debye/Angstrom or Debye/radian.
.....repeat this line NSX times (i.e. I = 1, NSX)
Print control
Printing in INTDER is controlled by the NPRT option. This is
a four digit number, DCBA, the meaning of which is as follows:
A = 0 ... default, standard output
>= 1 ... cubic and quartic force constants are printed
>= 2 ... the symmetrized B matrix is printed
>= 3 ... the A matrix (= B inverse) is printed
>= 4 ... the transpose of the symmetrized BB matrix is
printed
>= 5 ... linear transformation contributions to the force
constants are printed
B control of printing with the NTEST option
= 0 ... default, no printing of SR matrices
>= 1 ... analytic SR and Y matrices are printed as
governed by NTEST
>= 2 ... error matrices (SR analytic - SR numerical, and
perhaps Y analytic - Y numerical) are printed
as governed by NTEST
C control of printing with the NFREQ option
= 0 ... default, standard output
>= 1 ... the G matrix and its eigenvalues are printed if
NFREQ = 1, 3 or 4.
>= 2 ... the dipole moment derivatives with respect to
normal coordinates are printed if NFREQ does not
equal 0.
>= 3 ... eigenvectors for the zero frequencies in normal
coordinates are printed if NFREQ = 2 or 3.
equal 0.
D control of printing to the CHECK file
= 0 ... default, standard output
>= 1 ... messages from subroutines XIN, XOUT, YIN and
YOUT are suppressed
>= 2 ... force constants are printed in NINV = 2 format
>= 3 ... quadratic force constants are printed in the
format used by the BMATIN6 program
If NPRT is negative, the force constants will
also be written directly into the file called
BMAT in the appropriate place.
This is useful if one wishes to obtain an initial
force constant matrix at a low level of theory
for subsequent use in a high level optimization.
The force constants in BMAT are overwritten.
Thus, if it is a new BMAT file, it is important
to set it up with the appropriate number of
blank lines for the force constant matrix.
Also, make sure that the number of internal
coordinates used in INTDER and BMAT is the
same.
>= 4 ... input for use with the old GFMAT program is
printed
Dipole moment derivatives:
If NVEC = 1, then dipole moment derivatives are to be read from
FILE17 (if NINV = 0) or from FILE18 (NINV = 1).
The information required in FILE17 is:
1. FORMAT(5X,I5,3F20.10)
NA number of atoms
ICHG total charge on molecule
MUX X component of dipole moment
MUY Y component of dipole moment
MUZ Z component of dipole moment
2. FORMAT(3F20.10)
((U(I,J), J=1,NC), I=1,NC)
Cartesian dipole moment derivatives in Debye/A
(NC = 3*NA)
The information required in FILE18 is:
1. FORMAT(5X,I5,3F20.10)
NA number of atoms
ICHG total charge on molecule
MUX X component of dipole moment
MUY Y component of dipole moment
MUZ Z component of dipole moment
2. FORMAT(3F20.10)
((U(M,N), M=1,NSX), N=1,3)
internal (symmetry) coordinate dipole moment
derivatives in Debye/A or Debye/radian
(NSX = number of internal coordinates.
NSX = NSYM. If NSYM = 0 then NSX = NS.)
The layout is as follows:
Coord
d mux / d s 1 2 3
4 5 6
....
d muy / d s 1 2 3
4 5 6
....
d muz / d s 1 2 3
4 5 6
....
________________________________________________________________________
1 57. VIBLRG
------
A. VIBLRG takes a sequence of finite displacement first derivatives in
Cartesian coordinates and calculates the corresponding second
derivative matrix.
B. References: none
C. Files required: HVIB15
Temporary files used: none
Files generated: FILE6
FILE15
D. Input format:
FILE15
1. FORMAT(4X,I2)
ICEN ......... number of centers
2. FORMAT(A3,4X,A2,I1,A1,I3)
LBL
IATOM
INUM
ISIGN
NPAIR
3. FORMAT(F10.6)
DISP(J) ......... displacement in bohr
..... repeat this card INUM times
4. FORMAT(20(A2,I1,1X)
(SCRCOD(J),SCRNUM(J),J=1,ICEN)
5. FORMAT(20(A2,I1,1X)
(SCRCOD(J),SCRNUM(J),J=J1,J2)
..... repeat this card NPAIR times
6. FORMAT(3F10.8)
VECT(J),J=K,K1
7. FORMAT(3F10.8)
REF(K+1), REF(K+2), REF(K+3)
..... repeat this card ICEN times
8. If atom mass need to change
FORMAT(F20.12)
AMASS1 ......... changing mass for IATOM
9. FORMAT(2A1,2X,I1)
ISGN,LET,KNT
10. FORMAT(4F20.8)
FORCEP(J,I),J=MIN,MAX
..... repeat this card ??
11. FORMAT(4F20.8)
FORCEN(J,I),J=MIN,MAX
..... repeat this card ??
________________________________________________________________________
1 58. FORM15
------
A. FORM15 may be used as an alternative to the VIBLRG program.
A symmetry unique set of Cartesian coordinate gradients is used
(the molecular point group should be D2h or its subgroups).
The second derivatives are written in standard format to FILE15,
which can then be used with the NORMCO program to calculate the
vibrational frequencies.
B. References: none
C. Files required: INPUT
FILE11
FILE30
Temporary files used: none
Files generated: FILE15
D. Input format:
# FORM15 #
1. FORMAT(I5)
( FORMAT(A5,2I5) ) |
SYMTYP |
NDEG |
IPRNT
FILE11
The format is the standard FILE11 containing cumulated gradients.
... except you need to add an extra line (blank or something |
you want for printing in free field format) after each title. |
________________________________________________________________________
1 59. WRIT17
------
A. WRIT17 takes a sequence of finite displacement dipole moments in
Cartesian coordinates and calculates the corresponding dipole moment
derivative matrix. This is then written to FILE17.
B. References: none
C. Files required: TOTDIP
Temporary files used: none
Files generated: FILE17
D. Input format:
TOTDIP
1. FORMAT(I5)
NATOM ...... number of atoms
2. FORMAT(F20.10)
DIS ...... value of displacement
3. FORMAT(3F20.10)
X1 ...... X component of dipole moment
Y1 ...... Y component of dipole moment
Z1 ...... Z component of dipole moment
..... for positive displacement
4. FORMAT(3F20.10)
X2 ...... X component of dipole moment
Y2 ...... Y component of dipole moment
Z2 ...... Z component of dipole moment
..... for negative displacement
..... repeat lines 3 and 4 NATOM times
________________________________________________________________________
1 60. WRIT20
------
A. WRIT20 takes a sequence of finite displacement second derivatives in
Cartesian coordinates and calculates the corresponding third
derivative matrix. This is then written to FILE20.
B. References: none
C. Files required: INPUT (# FDGEOM #)
TOTAL15
Temporary files used: none
Files generated: FILE6
FILE20
D. Input format:
# FDGEOM # (default values used if # FDGEOM # is not found)
1. FORMAT(4I5)
ISYM30 = 0 ... (default)
I2PNTD = 2 ... (default)
IFLAG3 = 0 ... (default)
IPRINT = 0 ... (default)
2. FORMAT(*)
FINDIF = 0.0 ... (default)
TOTAL15
1. FORMAT(3I5)
NATOM ........ the sequence number of the atom being displaced
NAXIS = 1 ... positive displacement
= -1 ... negative displacement
NDISPS ........ total number of displacements
2. The second derivatives, exactly as they are printed in a
normal FILE15 (including the first line containing the total
number of atoms).
...... repeat lines 1 and 2 a total of 6 * NATOM times. (NDIPS may
be omitted from subsequent lines).
________________________________________________________________________
1 61. WRIT24
------
A. WRIT24 takes a sequence of finite displacement third derivatives in
Cartesian coordinates and calculates the corresponding fourth
derivative matrix. This is then written to FILE24.
B. References: none
C. Files required: INPUT
TOTAL20
Temporary files used: none
Files generated: FILE6
FILE24
D. Input format:
# FDGEOM # (default values used if # FDGEOM # is not found)
1. FORMAT(4I5)
ISYM30 = 0 ... (default)
I2PNTD = 2 ... (default)
IFLAG3 = 0 ... (default)
IPRINT = 0 ... (default)
2. FORMAT(*)
FINDIF = 0.0 ... (default)
TOTAL20
1. FORMAT(3I5)
NATOM ........ the sequence number of the atom being displaced
NAXIS = 1 ...
= -1 ...
NDISPS ........ number of displacement
2. The third derivatives, exactly as they are printed in a
normal FILE20 (including the first line containing the total
number of atoms).
...... repeat lines 1 and 2 a total of 6 * NATOM times. (NDIPS may
be omitted from subsequent lines).
________________________________________________________________________
1 62. INTDIF
------
A. INTDIF numerically calculates derivatives up to fifth order in
internal coordinates.
B. References:
C. Files required: INTDER1 (# INTDIF #)
FILE12A
FILE16A
FILE21A
Temporary files used: FILE91
FILE92
Files generated: CHECK
INTDIFO
IDER
D. Input format:
The input for this program is read from a file called INTDER1, which
has the same format as INPUT. The first section in this file should
be the # FILES ## input to tell the program what the temporary files
will be called. This is followed by:
# INTDIF #
1. FORMAT(A10)
LABEL = FCONSTDIF ... Perform a standard finite-difference
calculation to obtain force constants
up to fifth order with the use of
analytic derivatives up to
third order.
= DIATOM ... For the case of a diatomic molecule
(or if only one coordinate is
pertinent), energy points and/or
gradients are used to locate the
energy minimum and obtain force
constants up to fourth order.
The following is only used if LABEL = FCONSTDIF:
2. FORMAT(6I5)
NS number of (internal) coordinates
NDER highest order for which analytic derivatives are available.
First derivatives are read from FILE12A,
second derivatives are read from FILE16A, and
third derivatives are read from FILE21A.
For a description of the format required, see below.
NPERM number of symmetry operations needed to generate
all of the coordinates from the symmetry unique set.
NPAR number of coordinates for which only positive displacements
are needed.
NPRT = 0 ... normal printing
> 0 ... more output
NMORSE = 0 ... standard calculation
= 1 ... increased accuracy in the numerical diagonal force
constants is to be achieved by approximate methods.
3. If NPERM does not equal 0:
FORMAT(16I5)
(IPERM(I,J), J=0,NS)
IPERM(I,0) is the coordinate generated by the Ith symmetry
operation
IPERM(I,J) is the coordinate into which the Jth coordinate is sent
by the Ith symmetry operation. IPERM(I,J) can be
negative if a positive displacement for one coordinate
is mapped into a negative displacement for another
(perhaps the same) coordinate.
.....repeat line 3 NPERM times (i.e. I = 1, NPERM)
4. If NPAR does not equal 0:
FORMAT(I5)
NI is a coordinate for which only a positive
displacement is given (see line 5)
5. If NPAR does not equal 0:
FORMAT(16I5)
(IPAR(I,J), J=1,NS)
For each NI (see line 4), IPAR(I,J) contains the
parity (+1 or -1) of coordinate J under the
symmetry operation which generates the -NI
displacement from the +NI displacement.
.....repeat lines 4 and 5 NPAR times
6. If NMORSE = 1:
FORMAT(16I5)
(IMORSE(I), I=1,NS)
IMORSE(I) = 0 ... No special procedure used. The N-th order force
constants are obtained via central difference
formulas and (N-1)-th order analytic derivatives.
IMORSE(I) = 1 ... This value is appropriate for bond-stretching
coordinates. Either simple bond lengths or
normalized symmetry bond lengths are possible.
The effect is to assume a Morse oscillator to
reduce the numerical error in the finite-
difference diagonal force constants. (A
reduction of the error by a factor of 5
to 10 is typical.)
IMORSE(I) = 2 ... This value is appropriate for any
coordinate. For the N-th order diagonal
force constant, numerical values based on
both (N-1)-th and (N-2)-th analytic
derivatives are used to improve the
accuracy. (A reduction of the numerical
error by a factor of 10 to 100 is typical
for bond-stretching coordinates.)
The following is only used if LABEL = DIATOM:
2. FORMAT(3I5)
NPOINT number of points (>=3)
NDER highest order for which analytic derivatives are available
NGUESS = 0 ... normal run
= 1 ... If 3 =< NPOINT <= 4, then approximate (fixed)
values of the cubic and quartic force
constants (F3 and F4) can be used to
interpolate (or extrapolate) the equilibrium
geometry.
3. FORMAT(2F20.12)
S(I) internal coordinate S (in Angstrom)
E(I) corresponding energy (in hartrees)
.....repeat line 3 NPOINT times (i.e. I=1,NPOINT)
4. If NGUESS not = 0 and NPOINT = 3:
FORMAT(2F20.12)
F3 in mdyne/Angstrom**2
F4 in mdyne/Angstrom**3
5. If NGUESS not = 0 and NPOINT = 4:
FORMAT(F20.12)
F4 in mdyne/Angstrom**3
Information required in FILE12A:
Accumulated first derivatives in (symmetry) internal coordinates.
1. FORMAT(I5,F12.8,F20.10)
NI number of coordinate which is displaced in this
geometry (= 0 for reference geometry)
DELTA value of displacement (in Angstrom or radian)
ENERGY corresponding energy (in hartree)
2. F1(M) first derivatives as written to FILE12 by INTDER
.....repeat lines 1 and 2 for each displacement
Information required in FILE16A:
Accumulated second derivatives in (symmetry) internal coordinates.
1. FORMAT(I5,5X,F12.8)
NI number of coordinate which is displaced in this
geometry (= 0 for reference geometry)
DELTA value of displacement (in Angstrom or radian)
2. F2(M,N) second derivatives as written to FILE16 by INTDER
.....repeat lines 1 and 2 for each displacement
Information required in FILE21A:
Accumulated third derivatives in (symmetry) internal coordinates.
1. FORMAT(I5,5X,F12.8)
NI number of coordinate which is displaced in this
geometry (= 0 for reference geometry)
DELTA value of displacement (in Angstrom or radian)
2. F3(M,N,P) third derivatives as written to FILE21 by INTDER
.....repeat lines 1 and 2 for each displacement
________________________________________________________________________
1 63. ANHARM
------
A. ANHARM calculates anharmonic constants using the second-order
perturbation approach.
It also transforms SCF second, third and fourth derivatives from
Cartesian coordinates to normal coordinates and calculates a
variety of spectroscopic constants.
At the present, the anharmonic analysis can only be performed for
asymmetric top molecules.
B. References:
D. A. Clabo, W. D. Allen, R. B. Remington, Y. Yamaguchi and H. F.
Schaefer, Chem. Phys. 123 (1988) 187.
C. Files required: INPUT (# ANHARM #)
FILE15
FILE20
FILE24 (if ITHREE = 1)
FILE30 (if IFOUR = 1)
Temporary files used: none
Files generated: CHECK
FILE6
D. Input format:
# ANHARM #
1. FORMAT(10I5)
ISOTOP = 0 or 1 use regular atomic masses
= 2 or more number of isotopomers
IGEOMT = 0 ... read in geometry from FILE30
= 1 ... read in geometry from INPUT file
ITHREE = 0 ... do not read in third derivatives
= 1 ... read in third derivatives from FILE20
IFOUR = 0 ... do not read in fourth derivatives
= 1 ... read in fourth derivatives from FILE24
IFREQ = 0 ... calculate 3N-6 (or 3N-5) sets of anharmonic
constants
= 1 ... calculate IFREQ sets of anharmonic constants
ICORIO = 0 ... use default threshold (100.0) for Coriolis
resonance
= 1 ... read in threshold value
IFERM1 = 0 ... use default threshold (100.0) for Type 1
Coriolis resonance
= 1 ... read in threshold value
Type 1 is w(I) = w(J)
IFERM2 = 0 ... use default threshold (100.0) for Type 2
Coriolis resonance
= 1 ... read in threshold value
Type 2 is w(I) + w(J) = w(K)
ISIGMA = 0 ... the asymmetry parameter (sigma) needed for
the centrifugal distortion constants is
calculated using A(0'), B(0') and C(0')
constants
= 1 ... sigma is calculated using A(E), B(E) and C(E)
constants
IPRNT = 0 ... normal printing
= 1-3 .. more output
2. If IFREQ is not equal to 0:
FORMAT(10I5)
(NFRQ(I), I=1,IFREQ)
This option is typically used to rearrange
the frequencies to spectroscopic ordering.
3. If ICORIO is not equal to 0:
FORMAT(F20.10)
CLIMIT threshold value for Coriolis resonance
4. If IFERM1 is not equal to 0:
FORMAT(F20.10)
FLIM1 threshold value for Coriolis resonance Type 1
5. If IFERM2 is not equal to 0:
FORMAT(F20.10)
FLIM2 threshold value for Coriolis resonance Type 2
6. If IGEOMT is not equal to 0:
FORMAT(4F20.10)
COORD(1,I) x coordinate
COORD(2,I) y coordinate
COORD(3,I) z coordinate
W(I) atomic mass
.....repeat this line for each atom
7. If ISOTOP is greater than 1:
FORMAT(F20.10)
W(I) atomic mass
.....repeat this line for each atom
.....section 7. should occur ISOTOP-1 times. The first anharmonic
constant calculation is always performed with regular atomic
masses (unless IGEOMT = 1, in which case the masses from
section 2. above are used).
________________________________________________________________________
1 64. READ30
------
A. READ30 is a utility program used to read the binary FILE30 and
write out the information to FILE6 in human-readable form (almost).
B. References: none
C. Files required: INPUT
FILE30
Temporary files used: none
Files generated: CHECK
FILE6
D. Input required: none
________________________________________________________________________
1 65. SCFX |
---- |
|
A. SCFX solves the Hartree-Fock equations for excited state |
wavefunctions where the symmetry of the excited state is the same |
as the ground state. |
This program may also be used for TCSCF wavefunctions where the two |
special orbitals have the same symmetry. |
|
B. References: |
G. Fitzgerald and H. F. Schaefer, J. Chem. Phys. 83 (1985) 1162. |
W. D. Allen and H. F. Schaefer, J. Chem. Phys. 87 (1987) 7076. |
|
|
C. Files required: INPUT (# SCFEX ## and # TFOCK ##) |
FILE30 |
FILE34 |
|
Temporary files used: FILE92 |
|
Files updated: FILE30 MO coefficients |
|
Files generated: CHECK |
FILE6 |
FILE47 |
FILE48 |
FILE49 |
|
D. Input format: |
# SCFEX ## |
|
1. FORMAT(A80) |
ALABEL title for SCFEX output |
|
2. FORMAT(14I5) |
(1) IAVRQ IAVRQ=0,1,2,...,10 |
Damp the orbital variations in the open shell |
symmetry block. LAMBDA=IAVRQ/10. LAMBDA=1 means |
to take the full step while LAMBDA=0 gives no |About
The original Psi 1.0 release from the 1980s.
Resources
Stars
Watchers
Forks
Releases
No releases published
Packages 0
No packages published